PROCEEDINGS OF THE Academy of Natural Sciences OF PHILADELPHIA Volume lv 1903 philadelphia : The academy of Natural Sciences LOGAN SQUARE 1903-1904 The Academy of Natural Sciences of Philadelphia, February 5, 1904. I hereijy certify that printed copies of the Pkoceedixos for 1903 have bc>r>ii mailed as follows: Pa^es 1- 32 mailed April 33- SO Sl-128 • " 129-160 161-20S 209-256 257-336 337-352 353-400 401-528 529-576 577-624 625-640 ' 641-688 689-704 705-752 753-784 7S5-816 April 10. 1903 Ma^- 7, 1903 Mav 16. 1903 Mav 22, 1903 Junf 4. 1903 June 30, 1903. July 8, 1903 July 10, 1903 August 18. 1903 September 4, 1903 September 29. 1903 October 6, 1903 November 10. 1903 November 25. 1903 December 4, 1903 December 16. 1903 Januarv 15. 1904 February 1, 1904 EDWARD J. NOLAN, Recording Secretary. COMMITTEE ON PUBLICATION: Heniiy Skinner, M,D., Philip P. Calvert. Ph.D.. Henry A. Pilsbry, Sc.D., Wither Stone, Edward J. Nolan, M.D., The President, Samuel G. Dixon, ]\I.D., ex-officio. Editor: Edward J. Nolan. M.D. 7(15" CONTEXTS. For Announcei7ients , RejMrts, etc., see General Index. Anderson. ^ifALComi P., and Joseph Grixnell. Birds of the Siskiyou Mountains, California : a Problem in Distribu- tion, 4 Banks. Xathax. A List of Arachnida from Hayti, with De- scriptions of New Species (Plate XV) 340 BiLGRAM. Hugo. Inclusions in Quartz, 700 Brown, Arthur Erwin. The ^'a^iations of Euta:'nia in the Pacific Subregi on, 286 Texas Reptiles and Their Faunal Relations 543 Calvert, P. P., Ph.D. Ganglia of Odonata, 760 Casey, Thomas L. Notes on the Conrad Collection of Vicksburg Fossils, with Descriptions of New Species, 261 Chamberlaix, Ralph V, Myi-iopods from Beulah, New Mexico, 35 New Lithobil from California and Oregon, 152 Chapman, Henry C, M.D. On a Collection of Anthropoids, . 719 CocKERELL, T. D. A. Bees of the Genus Nomada from Califor- nia 559 North Amei-ican Bees of the Genus Nomada, 580 Variations in the Snail-Genus Ashmunella, 615 Conklin, Edwin G., Ph.D. The Cause of Inverse Symmetr}-, . 753 Crawley, Howard. List of Polycystid Gregarines of the United States (Plates I, II, HI), 41 Nosema geophili, sp. n., a Myxosporidian Parasite of Geo]ihi- lus, ...... ^ 337 The Polycystid Gregarines of the United States (Plate XXX), 632 EiGENMAXX, Carl H., and Clarence Hamilton Kennedy. On a Collection of Fishes from Paraguay, with a Synopsis of the American Genera of Cichlids, 497 Fielde, Adele M. Supplementary Notes on an An I, .... 491 Experiments with Ants Induced to Swin:i 617 i 11 FoAVLER, Henry AV. Descri])tions of Several Fishes from Zanzi- bar Island, Two of which are New (Plates VI, VII, VIII), 161 Life Colors of Pcecilia limantouri and Description of a New Heros from Mexico, 320 Description of a New Gurnard from Florida, with Notes on the Colors of some other Florida Fishes, 328 Notes on a Few Fishes from the Mosquito Coast of Nicaragua 346 Descriptions of New, Little Known and Typical Atherinidse (Plates XLI-XLIV), 727 New and Little Known Mugilidte and Sphyra?nidte (Plates XLV-XLVI), 743 Description of a New Lantern Fish, 754 Frazer, Persifor, D. es-Sc. History of the Caribbean Islands from a Petrographic Point of View (Abstract), . . . 396 Harshberger, John W. The Mutation of Hibiscus moscheutos, 326 The Form and Structure of the Mycodomatia of Myrica cerifera (Plates XVI, XVII), ' 352 Keeley, Frank J, Preparation of Diatoms, 2 Inclusions in Quartz, 700 Monks, Sarah P. Regeneration of the Body of a Starfish, . . 35 Montgomery, Thomas H., Jr., Ph.D. Studies of the Habits of Spiders, particularly those of the Mating Period (Plates IV, V), 59 On the Morphology of the Rotatorian Family Flosculariidse (Plates XVIII, XIX, XX, XXI), . . \ 363 Supplementary Notes on Spiders of the Genera Lycosa, Pardosa, Pirata and Dolomedes from the Northeastern United States (Plate XXIX), 645 Moore, J. Percy. Polychreta from the Coastal Slope of Japan and from Kamchatka and Bering Sea (Plates XXIII, XXIV, XXV, XXVI, XXVII), 401 Descriptions of Two New Species of Polychteta from Wood's Hole, Massachusetts (Plate XL), 720 Some Pelagic Polychteta New to the Wood's PIolc Faima (Plate' LV), ' 793 PiLSBRY, Henry A., D.Sc. Mollusca of Western Arkansas and adjacent States, with a Revision of Paravitrea (Plates IX, X, XI), 193 Additions to the Japanese Land Snail Famia, No. VIII (Plate XIV) Ill PiLSBRY, Henry A., D.Sc. A New Japanese Pleiirotomaria (Plate XXII), 496 A New American Genus of Arionidre (Plate XXVIII), . . 626 Mexican Land and Freshwater Mollusks (Plates XI.NII- LIV), 761 A New Hawaiian Limnaea, 790 Rehn, James A. G. A Revision of the ( rtho]Dteroiis Genus Homoeogamia, 177 Studies in American Forficulidae, 299 Studies in Old World Mantidse (Orthoptera), 701 Rehn, James A. G., and T. T>. A. Cockerell. A New Genus of Stenopelmatinae (Orthoptera) from Mexico, .... 630 Skinner, Henry, M.D. A New Species of Nyctaleniou (Plate XIII), 298 Stone, Witmer. A Collection of Reptiles and Batrachians from Arkansas, Indian Territory and Western Texas, . . . 538 Racial Variation in Plants and Animals, with Special Refer- ence to the Violets of Philadelphia and Vicinity (Plates XXXI-XXXIX), 656 Stone, Witmer, and James A. G. Rehn. On the Terrestrial Vertebrates of Portions of Southern New Mexico and Western Texas, 16 True, Frederick W. A Note on the Common Bottle-nosed Por- poise of the North Atlantic, Tursiops truncatus (Mon- ■ tagu), 313 \''anatta, E. G. a List of Shells Collected in Western Florida and Horn Island, Mississippi, 756 Wheeler. William Morton. A Revision of the North Ameri- can Ants of the Genus Leptothorax Mayr (Plate Xll). 215 PROCEEDINGS OF THE ACADEMY OF NATURAL SCIENCES OF PHILADELPHIA. 1903. January 6. Mr. Arthur Erwin Brown, Vice-President, in the Chair. Twenty-one persons present. The Council reported that the following Standing Committees had been appointed to serve during the ensuing year : On Library. — Dr. C. N. Peirce, Thomas A. Robinson, Henry C. Chapman, M.D., Charles Schaeffer, M.D., and George Vaux, Jr. On Publications. — Henry Skinner, M.D., Henry A, Pilsbry, Sc.D., Phihp P. Calvert, Ph.D., Witmer Stone, and Edward J. Nolan, M.D. On Instruction and Lectures. — Benjamin Smith Lyman, Thomas H. Montgomery, Ph.D., Henry A. Pilsbry, Sc.D., Charles Morris, and PhiUp P. Calvert, Ph.D. On Finance. — Isaac J. Wistar, Wilham Sellers, John Cadwalader, Edwin S. Dixon, and the Treasurer. Council's Committee on By-Laws. — Isaac J. Wistar, Theodore D. Rand, Arthur Erwin Brown, and Thomas H. Fenton, M.D. The President is, ex-officio, a member of all Standing Committees. Mr. George Vaux, Jr., was appointed the Solicitor of the Academy. proceedings of the academy of [jan., January 13, Mr. Arthur Erwin Brown, Vice-President, in the Chair. Twelve persons present. The death of Bushrod W. James, M.D., a member, was announced. January 20. Mr. Arthur Erwin Brown, Vice-President, in the Chair. Sixteen persons present. Papers under the following titles were presented for pubhcation : "Myriapoda from Beulah, N. M./' by Ralph V. Chamberlain. ''List of the Polycestid Gregarines of the United States," by Howard Crawley. Preparation of Diatoms.— Mr. Frank J.Keeley called attention to a method recently employed in studying the structure of diatoms, which is supplementary to those previously described by him :— Mountino- broken valves on edge so as to obtain a sectional view, and mounting in a manner that permitted of successively immersing the forms in media of varying refraction indices. ■,• ci The present method consists in depositing on the diatoms a thm tiim of silver, using the solution ordinarily employed for silvering mirrors, which if dropped on the cover-glass containing the diatoms, will silver the latter to a considerable extent before any appreciable quantity of the metal i^ deposited on the glass. The finer the irregularities on the sur- face the heavier will be the deposit of silver, and the best results are obtained on gatherings containing broken and corroded forms, sepa- rated plates "etc., the structure of which becomes quite apparent. Thus conclusions can be drawn as to the correct nature of a complete and perfect valve which will be more accurate than when it is itself examined, particularly if the markings are fine, when they become filled with silver. . ■ , . i Under favorable conditions, after mounting m balsam and examining bv transmitted light, valves or portions of them may be found having the appearance of plates of perforated metal, and while the results so far have principally been corroborative of those obtained by previous methods some features not distinguishable under ordinary conditions are rendered apparent, among which may be mentioned a ring of pro- cesses near the margin of the valve of Coscinodiscus suhfilis, which extends toward the interior of the frustule. , ■ -^r . . ' The character of the raphe is particularly weU displayed in ^av^c^da and its allies, as well as the small channels in the central and terminal nodules, which doubtless have some function m connection with tlie 1903.] NATURAL SCIENCES OF PHILADELPHIA. 3 protoplasmic currents causing the motion of diatoms. In Navicula rhomboides it becomes apparent that the raphe is single as in other species of this genus, and that the double raphe, based on which this form has been placed in a separate genus, is an optical illusion. This process differs from that by which the late A. Y. Moore plated diatoms, as his were covered with a hea\'y layer of silver or gold and examined as opaque objects. It is rather a staining process, but like- wise differs from the methods of Moore and others in this direction, which filled the cavities in the valves with opaque matter, while in this case the silica itself is rendered opaque, or nearly so. As in all microscopical investigation, it is necessary to use judgment in interpreting appearances seen, as they will vary materially in the same species, depending on whether the entire structure is preserved. Isthmia affords an excellent example. Some valves may be found in which the coarse markings appear as holes in the blackened sihca, showing that the secondary structure is entirely gone. Others wall show the secondary structure in a similar manner as holes in a darkened plate, but still others will have the large oval markings entirely covered with a film of silver, indicating that there is a still finer and ordinarily invisible "tertiary structure'' or plate, which serves as a basis on which the silver is deposited. Exactly the same conditions may be noted on the external plate of Coscinodiscus asteromphalus , but there is never an indication that the eye-spots on the inner plate are anything but clear openings. January 27. The President, Samuel G. Dixon, M.D., in the Chair. Twenty-two persons present. The following were elected members : James Spear, Henry A. Lang, and Edw. B. Meigs. William Morton Wheeler, of Austin, Tex., and Theodore Boveri, of Wiirzburg, were elected correspondents. The following were ordered to be printed : PROCEEDINGS OF THE ACADEMY OF [Jan., BIRDS OF THE SISKIYOU MOUNTAINS, CALIFORNIA: A PROBLEM IN DISTRIBUTION. BY MALCOLM P. ANDERSON AND JOSEPH GRINNELL. The birds of the following list were taken or observed by Malcolm P. Anderson in the extreme northwestern part of California, between Sep- tember 6, 1901, and March 10, 1902. His field-notes form the basis of the present paper, while Joseph Grinnell is responsible for the critical remarks on specimens and distribution. The region explored occupies the northwestern corner of Siskiyou county, close to the Oregon boundary. Most of the observations per- tain to the region of Horse and Seiad creeks, which head in the Siskiyou range. Horse creek flows into the Klamath river from the north about four miles below the old mining camp called Oak Bar, and about nine miles above the mouth of Scott river. Seiad creek joins the Klamath near Seiad Post Office, which is eleven miles below Scott river, and therefore twenty miles from Horse creek. Although these two streams empty so far apart they head near together, there being but one ridge between the main branch of each. Oak Bar is at an approximate altitude of 2,800 feet, while the crest of the Siskiyou Mountains, some seventeen miles north of the Klamath river, is 7,000 feet in elevation. The highest points of the range in the neighborhood of Horse creek approach 8,000 feet. West of Seiad creek the country changes abruptly, becoming far rougher, with the peaks in some instances rising higher. Near the mouths of both Horse and Seiad creeks the canyons expand into valleys, which are occupied l^y a few farmer-miners. The valley of Seiad creek is much the larger and is known as Seiad Valley. After the first of November camp was made about ten miles up Horse creek, at a little farm called Grater's ranch. The region under consideration is covered with an open forest ex- tending from the Klamath up to within a few hundred feet of the summit of the range. Here the trees become scattering and in places give way to large stretches of chaparral, with here and there mead- ows in which grass grows waist high unless disturbed by cattle. The Life Zones represented in the region seem to be : Transition, which extends from the Klamath up to about the 6,000-foot contour line, and Canadian, which caps the range, with perhaps dilute Hudsonian in 1903.] NATURAL SCIENCES OF PHILADELPHIA. 5 places. The following trees identified will give a better idea of the con- ditions among which the birds were found : Trees of the Transition Zone. Pseudotsuga tnua'onata. Douglas Spruce. This is the most conspicuous tree of the region. On Horse creek it is found from the Klamath up to an altitude of 5,500 to 6,000 feet. Libocednis deciirrens. Incense Cedar. This was also a very common tree from the Ivlamath up to about 5,000 feet. Pinus lamhertiana. Sugar Pine. This is the largest tree of the region, and is common along the sum- mits of White Cloud and Johnny O'Neil ridges from the Klamath up. It is most abundant and of greatest size in the basin of Salt creek about a mile from Grater's. Not observed above 4,500 feet. Pinus ponderosa. Yellow Pine. Seen from the Klamath up to about 4,000 feet; frequently of large size. Pinus attenuata. Scrub Pine. This tree grows commonly in the upper part of Seiad Valley, and was noticed along Seiad creek up to 4,000 feet. Quercus calif ornicus. Black Oak. Though there are several species of oaks in the region, the only one that is very conspicuous is the California black oak. It is found from the KJamath to an elevation of about 4,000 feet. In several places it forms groves of considerable size to the exclusion of coniferous trees. Arbutus menziesi. Madrone. Seen frequently among the oaks up to their limit, Alnus sp.? A species of alder grows along all the streams. Acer macrophyllum. Big Leaf Maple. Observed along streams, but limits not noted. ' ' Trees of the Canadian Zone. Abies magnifica shastensis. Shasta Fir. This tree mixes with the following in the upper edge of Transition, and extends throughout the Canadian zone. 6 PROCEEDINGS OF THE ACADEMY OF [Jan., Abies concolor lowiana. A fir probably referable to this form is found in the lower portion of Canadian and higher Transition, where it occurs in company with the Douglas spruce. In the canyon of Horse creek it was seen down to an altitude of 3,100 feet. Pinus monticola. Mountain Pine. This tree is frequent in the Canadian zone, to which it seems restricted. Hudson: AN Zone. Tsuga mertensiana. Black Alpine Hemlock. Hemlocks grow in abundance near the head of the east fork of Horse creek and elsewhere along the range in cool places. During a tramp along the crest this tree was seen repeatedly. Notes on the Birds of the Region. Ardea herodias herodias. Great Blue Heron. On January 25 a great blue heron was seen on Seiad creek, about four miles above the mouth. Again, on March 2, two herons were observed on the Klamath river, between Oak Bar and Walker Post Office. Oreortyx pictus plumiferus. Mountain Partricige. On September 12 a small boy brought in two cjuail which had been shot near the house on Grater's Ranch. Subsequently the species was detected only once : a flock of six or eight were seen in Seaid Valley on December 20. The two September birds were preserved, and in spite of extremely abraded plumage, are plainly referable to plumiferus, and not pictus, as one might expect. One of the specimens is adult and shows a few new feathers appearing on the scapulars, breast and sides, but there is no evidence of approach toward the brown-necked humid coast form. This is an interesting example of the westward extension of a Sierra Nevadan form. Bonasa umbellus sabinei, Oroyon Ruffed Grouse. The people of the Siskiyou state that this l^ird is invariably seen on the ground or on fallen logs. The observations of the past winter go to confirm this, for all the grouse seen were upon the ground or flying near it, and in all cases were in groves of small Douglas spruce, such as grow near or in the canyon bottoms. There is another grouse known to the inhabitants of this region. This l^ird is said to l)e found mainly in trees and is doubtless the sooty grouse (Dendragapus obscurus fuligi- nosus). In the latter part of Februar}- and early in IMarch, the drum- ming of what was thought to be this latter species was heard, but not even a glimpse was obtained of the performer. Foiu' specimens of the 1903.] NATURAL SCIENCES OF PHILADELPHIA. 7 Oregon ruffed grouse were preserved, and the species may be considered fairly common through the winter in this region. This bird has been found previously inCaUfornia only in the redwood forests in the vicinity of Humboldt Bay, so that the present record extends its known range considerably to the eastward. Douglas, in the original description of his "Tetrao Sabini," gives its range as "from Cape Mendocino" north, etc. But recent records, up to the present, indicate ordinarily a dense forest habitat. Buteo borealis calurus. Western Red-tailed Hawk. An immature male red-tail was shbt on October 11 on the very sum- mit of the Siskiyou, where the Oregon trail crosses. Falco sparverius phalsena. Western Sparrow Hawk. One was observed on March 2 between Oak Bar and Walker on the Klamath. Ceryle alcyou. Belted Kingfisher. A single individual was observed flying up Horse creek on January 1 6. Dryobates villosus harrisi. Harris Woodpecker. This bird is very common in the groves of black oak which are located here and there on the lower parts of the Siskiyou range. It is also found in the coniferous forests, though in smaller numbers. From October 24 to February 15 fifteen specimens were taken. These are variously intermediate in coloration between hyloscopus and harrisi, but average nearest the latter. Five of the birds have all the white areas strongly suffused with smoke-gray or brown, nearly but not quite as deeply as in specimens from western Oregon and Washington. Four have all the areas practically pure white, while the other six are fairly intermediate between these two types. The whole fifteen present an uninterrupted series, from the whitest to the dingiest. The usual size of all the specimens, however, is that of harrisi; that is, decidedly larger than the white-breasted hyloscopus, which occupies nearly the whole of California in favorable localities outside of the extreme north- ern humid coast belt. Dryobates pubescens gairdneri. Gairdner Woodpecker. The Gairdner woodpecker is usually to be found in company with the flocks of mountain chickadees which frequent the black oak groves all winter. The oaks are their favorite working places, but they are also to be seen among the pines and spruces. The six specimens brought home are all quite near gairdneri. The smokiness of the lower surface is not so intense as in skins from western Oregon, but the size, especially of the feet, is decidedly that of the northwest coast form. PROCEEDINGS OF THE ACADEMY OF [Jan., Gairdneri occurs in California only in the extreme northwestern corner of the State, the greater portion west of the Sierras being occupied by D. p. turati. (See Fisher, Condor, IV, May, 1902, 68.) Xenopicus albolarvatus. White-headed Woodpecker. Only three birds of this species were seen in the Siskiyou region. All were shot from sugar pines in the neighborhood of Salt creek. They were obtained on October 10, January 10 and February 19, respectively. These specimens are exactly like the small-billed Sierra Nevadan birds. The known range of the species is thus materially extended, and here seems to be its westernmost station. Sphyrapicus varius ruber. Red-breasted Sapsucker. This woodpecker was not infrequently seen along Horse and Seiad creeks, mostly in the spruce forest at about 3,200 feet elevation. While not the extreme manifestation of ruber, the four specimens secured, by their large size and dark coloration, easily belong under this head. The white markings of wings and tail are much more restricted than in the Sierra Nevadan daggetti. Sphyrapicus thyroideus. Williamson Sapsucker. A single individual was found on a ridge above camp on December 19. It was at work in a tall Douglas spruce, and continued tapping hghtly until closely approached. This proved to be a female, and is darker colored than Sierran specimens compared with it. The head is browner, while the deeper yellow belly and extreme length of wing point toward a geographical variation perhaps parallel to S. v. ruber. Ceophloeus pileatus abieticola. Northern Kleated Woodpecker. The pileated woodpecker is one of the most conspicuous birds of the region. Its loud screech-hke call can be heard for many hundred yards across the canyons. Though often seen it is very wary and difficult of approach. Its favorite resorts seem to be the black oaks, but occa- sional birds are seen to alight in coniferous trees. One specimen had the stomach distended with large black ants, which it was gathering from an oak when shot. As shown by the following measurements (in miUimeters), the four specimens procured belong to the large northern race. There appear to be no significant differences between these and skins from the north Atlantic slope. Collector's No. Date. Wing. Tail. Tarsus. Bill from Nostril. d 56 9 63 9 141 9 156 Oct. 9 ' 13 Dec. 19 " 27 238 226 225 226 187 179 175 180 35 36 37 35 41 40 36 38 1903.] NATURAL SCIENCES OF PHILADELPHIA. 9 Melanerpes torquatus. Lewis Woodpecker. The first bird noticed on crossing the Klamath and entering Horse creek canyon was the Lewis woodpecker. The species was then, Sep- tember 6, abundant in oaks bordering the meadows near the mouth of the creek. On September 13, two specimens were shot from a spruce near Grater's farmhouse. There were a number of others about at the time, but at the report of the gun they made off down the canyon, and nothing was afterward seen of the species anywhere in the region. It probably occurs only as a migrant. Colaptes cafer saturatior. Northwestern Flicker. Flickers were seen in the yellow pine forests on the ridge west of camp all through the fall and early part of winter, but they were so excessively shy that none were secured until January 20. They were always abundant near the mouth of Seiad creek, in some open meadows lined with oaks, and alders overgrown with wild grapevines. As com- pared with specimens from the vicinity of Monterey, the type locality of coUaris, the three skins from the Siskiyou country show somewhat greater dimensions and also darker colors. This is especially noticeable on the back, which is a decidedly richer brown, the ground color being deep Prout brown. Sayornis saya. Say Phoebe. On September 16 a Say phcebe was shot from its perch on the barn at Grater's place. This was the only bird noted, and so is probably repre- sented in the region only as a migrant. Cyanocitta stelleri carbonacea. Coast Jay. This was the only jay detected in the Siskiyou Mountains, and was by no means as conspicuous as jays often make themselves. Only occasionally was one heard scolding in some cluster of spruces, or dis- covered hopping quietly from limb to limb in a black oak. Eight skins were obtained during the winter, and all proved to be quite typical of the humid coast race carbonacea (see Fisher, Condor, lY, March, 1902, 41). But a female specimen taken at Walker Post Office, Siskiyou county, March 11, 1902, is very different, approximating C. s. frontalis, though not as pale as the average of southern California birds. Walker is only about twenty miles due east of Seiad Valley, and about forty- five miles northwest of ]\It. Shasta. If this specimen represents the resident race of its locality, it furnishes a very interesting example of the possible close invasion of two races toward each other from their original separate centres of differentiation (until the area of intergrada- tion may ultimately disappear). 10 PROCEEDINGS OF THE ACADEMY OF [Jan.. Corvus americanus hesperis. California Crow. On October 1 a large flock of crows were noted on a meadow near the mouth of Horse creek. Later in October a single individual was seen in a meadow at Grater's. Numbers were observed along the Klamath near Oak Bar toward the middle of March (6th and 11th). Nucifraga columbiana. Clarke Nutcracker. Two individuals were observed on December 29 on the ridge between Horse and Salt creeks. Mr. Doney, a miner of more than ordinary in- teUigence and veracity, stated that this bird is common on the highest parts of the Siskiyou all through the summer. This is another indica- tion of the presence of a Hudsonian zone on these mountains. Carpodacus purpureas californicus. California Purple Finch. A male specimen was secured on February 1 on the ridge between Horse and Salt creeks. The bird was at the top of a large oak singing loudly. We took the singing as a sign of spring coming, but we had severer weather after that than before. Astragalinus psaltria psaltria. Arkansas Goldfinch. Soon after reaching Grater's ranch, September 7, these birds were common in a pasture near the house. But within a few days, and before specimens were secured, they took their departure. Junco hyemalis hyemalis. Slate-coloredJunco. A single male specimen was shot on January 25 from a black oak in the dooryard at Grater's. There were other j uncos about at the time, but all seemed to be of the Oregon form. This skin measures, wing 78 mm., tail 71. The tips of the dorsal feathers are shghtly brownish, but the bird agrees explicitly with selected north-Alaskan and Atlantic coast specimens. This makes about the twenty-second instance of the occurrence of this form in California. Junco hyemalis oreganus. Oregon Junco. i Junco hyemalis thurberi. Sierra Junco. J Juncos seemed to constitute the most common species throughout the Transition of the Siskiyou region. Not a day passed without num- bers being seen. On and after December 4, when the first snow fell, they became very common about camp, where they worked continu- ally in the old garden. Out of sixteen skins brought home, six are at a glance recognizable as oreganus; they are just like Seattle and Sitka specimens. Four others are just as typically thurberi when compared with Mt. Wilson skins, while the remaining six are partly immature and are less precisely referable. It might be surmised that the Oregon 1903.] NATURAL SCIENCES OF PHILADELPHIA. 11 juncos are winter visitants, while the thurberi type is the resident form, bnt this is pure conjecture. Melospiza cinerea merrilli. Merrill Song Sparrow. Two specimens of this song sparrow were taken on October 30 and December 14 respectively, from a tangle of blackberry briers near camp. These are much paler than morphna and phcea, with decidedly blackish streaks both on back and breast, thus approaching montana closely. It seems probable that it is this form that breeds in the region. Collector's No. Wing. Tail. 9 130 9 70 66 65 70 71 Melospiza cinerea morphna. Rusty Song Sparrow. A bird of this race was taken on November 6 from the tangle of briers in which the Merrill song sparrows were found. Another was taken later from a band of sparrows in Seiad Valley. The first of these (No. 78) inclines in paleness of hind neck and back toward merrilli. The other is nearer phcea in having a sooty streaking. But the large size in both cases places them rather in the morphna category. Collector's No. Wing. Tail. d" 110 d 78 71 69 76 77 Passerella iliaoa unalaschcensis. Shumagin Fox Sparrow. On October 9 two fox sparrows were taken from the same thicket mentioned in connection with the foregoing. These seem to be nearest unalaschcensis, though inclining toward meruloides in brownness of coloration. They are too small for irisularis. Making due allowance for wear, they are counterparts of summer specimens at hand from Prince William Sound, Alaska, which is about half-way between the metropolis of meruloides (Yakutat Bay) and that of unalaschcensis ( Aliaska Peninsula). Piranga ludoviciana. Western Tanager. A single specimen was shot from a sugar pine on White-Cloud ridge on September 12, evidently in migration. Vireo solitarius cassini. Cassin Vireo. The only specimen observed was secured September 12 from some alders bordering a small stream on White-Cloud ridge, doubtless in migration. 12 PROCEEDINGS OF THE ACADEMY OF [Jan., Vireo huttoni huttoni. Hutton Vireo. Hut ton vireos were found frequenting that part of the Transition zone in which oaks are most numerous, but were also often seen feeding in spruces. They remained throughout the winter in uniform numbers. There seems to be absohitely no difference between the six Siski- you skins and a series from the Santa Cruz Mountains and vicinity of Monterey, the latter being the type locality of huttoni. Monterey is in the southern extremity of the humid coast Transition, and there has always seemed a doubt as to the existence of a race obscurus in the northern part of the same belt. S. N. Rhoads, in the Auk, Vol. X, July, 1S93, p. 238, clearly pointed out the untenability of obscurus, but no one seemed to pay any attention to his remarks. As to the exist- ence of a really distinct form on Vancouver Island (F. h. insularis Rhoads), we have as j^et no specimens to indicate it. Looking at Anthony's description of obscurus, we find that he used southern California birds in comparison with his Oregon ones. Some skins now at hand from Pasadena, in moderately worn spring plumage, are not so green dorsally and laterally as birds of the same season from the Santa Cruz district. This is evidently due to the greater rate of fading and abrasion in the arid southern climate, for the majority of specimens in fresh fall plumage are indistinguishable, and equivalent extremes are to be picked out of both series. Sixty speci- mens have been examined in this connection. Dendroica auduboni. Audubon Warbler. A male was shot from a small flock on the west side of White-Cloud ridge on September 12. After that none were seen. Cinclus mexicanus. American Dipper. Fairly common from the Klamath up Horse and Seiad creeks to an altitude of 5,000 feet. Five skins were preserved. Troglodytes hiemalis pacificus. Western Winter Wren. This proved to be a common bird of the Siskiyou region, where it was almost always to be found in the blackberry tangles along the smaller streams. Seventeen skins secured serve only to emphasize the geo- graphical uniformity of this race in its long north-and-south range. CertMa americana zelotes. Sierra Creeper. Creepers are comparatively numerous in the region of Horse creek. They were often seen busily searching about the bases of the Douglas spruce or hurriedly climbing spirally up the trunk of some pine or oak. The eleven specimens procured do not seem to differ in the 1903.] NATURAL SCIENCES OF PHILADELPHIA. 13 least from Sierran birds. Nor do they look much different from Sitkan skins. However, we have no certified examples of occidentalis, which is a supposedly distinct form said to occupy the humid northwest coast belt. Sitta carolinensis aculeata. Slender-billed Nuthatch. This bird was seen everywhere from the Klamath river to the upper edge of Transition. It seems to lead the mixed flocks of Parus, Regulus, etc., with which it is nearly always seen. Seventeen specimens were obtained. Sitta canadensis. Red-breasted Nuthatch. The Canada nuthatch was far less conspicuous than the preceding species, though it was probably no less common. It also accompanied the mixed flocks of chickadees and kinglets to some extent, but seemed to prefer the coniferous trees to the oaks which the slender-billed nut- hatch so largely frequents. Previous to December 26 no individuals were seen. The specimen shot on that date was rather high up in Transition, and was, it seems probable, just descending from the Cana- dian zone. For the species was shortly afterward common. Parus gambeli. Mountain Chiclcadee. This chickadee was constantly present among both the oaks and coniferous trees throughout Transiton and up into lower Canadian. The fourteen specimens brought home have on an average smaller bills than a series from the mountains of southern California (just as with Xenopicus). This is the westernmost record for this species. Parus rufescens rufescens. Chestnut-sided Chickadee. This bird was almost always found in company with the mountain chickadee, forming mixed roving bands. The series of twenty-one specimens is uniform and quite like birds from Sitka and Oregon. It would be highly important to know whether this species and the last occur together in summer also, and breed in the same locality. Psaltriparus minimus. California Bush tit. This was another very common bird in the black oak forests, but occurred in the coniferous woods as well. It was not, however, observed above the limit of Pinus ponderosa. The ten specimens taken present no differences from corresponding plumages of birds from cen- tral and southern California. Nor are there any characters to dis- criminate the few specimens we have from Oregon and Washington from those from California. Judging from the specimens at hand, there is little or no evidence of a race californicus. There is considerable 14 PROCEEDINGS OF THE ACADEMY OF [Jan., variation in shade of back and pileum due to wear and fading, but the fresh fall birds all look alike. Kegulus satrapa olivaceus- Western Golden-crowned Kinglet. This bird, when observed at all, was always in the flocks of chick- adees and nuthatches. Seven specimens, all males, were secured during the winter. HylocicMa guttata nana. Dwarf Hermit Thrush. On December 20 the only small thrush seen was shot from an alder in Seiad Valley. This specimen, on account of its small size and brown coloration, is easily referable to the northern humid coast form nana. Merula migratoria propinqua. Western Robin. Seen occasionally throughout the winter, almost always with the varied thrushes. It ranges from the Klamath river up to the limit of the black oak timber at least. Ixoreus nsevius naevius- Varied Thrush. The varied thrush, although seen throughout the winter, was not common until January, when large flocks appeared on the meadows in Seiad Valley. (This valley is almost free from snow). When the snow left the meadow at Grater's during February, numbers of these thrushes appeared there. At the same time they became common in the oak forests. Sixteen specimens were preserved. These average in characters close to the Sitkan form. Sialia mexicana occidentalis. Western Bluebird. A male bluebird was shot from a small flock on White-Cloud ridge, September 12. After this none were seen until February 9, when a bird appeared near camp. They soon after became common both in the vicinity of Grater's and in Seiad Valley. Summary. The foregoing list is a remarkable one as showing the association of a number of birds not usually found together. The Siskiyou Mountains are evidently on the narrow line of mergence between the humid coast fauna and the arid Sierran fauna. Representative species of the two areas are here found in about equal numbers, in winter at least, as shown by the following comparative table : 1903.] NATURAL SCIENCES OF PHILADELPHIA. 15 Humid Coast Fauna. Arid Sierran Fauna. Bonasa umbellus sabinei, Dryobates viUosus harrisi, Dryobates pubescens gairdneri, Sphyrapicus varius ruber, Colaptes cafer saturatior, Cyanocitta stelleri carbonacea, Dendragapus obscurus fuliginosus, Oreortyx pictus plumiferus, Xenopicus albolarvatus, Sphyrapicus thyroideus, Nucifraga columbiana, Certhia americana zelotes, (Junco hyemalis oreganus,) {Melospiza cinerea jnorphna,) Parus rufescens rufescens. Junco hyemalis thurberi, Melospiza cinerea inerrilli, Parus gambeli. Jn the case of species represented in both areas but by distinct sub- species, sometimes it is the humid form that is present and sometimes the arid form (in a few cases there are intermediates). For instance, we find the typical coast form of Cyanocitta, but the typical arid race of Oreortyx. And in the case of distinct species, as a result of eastward and westward invasion respectively, the ranges of the separate repre- sentatives here overlap; for example, Parus rufescens and Parus gam- beli. We find that as a rule two faunae (of the same zone) are occupied each by a different subspecies of the same species, or a different species of the same genus ; and that a single genus is seldom represented in one fauna by more than one species. (If it is, then it seems to have come through the invasion of one of the congeneric forms from an adjacent fauna, as with the two chickadees just mentioned.) This seems to me a strong argument in favor of the theory that isolation (either by long distance or intervention of barriers) has been an absolutely essential condition to the differentiation of species. A fauna is a certain assem- blage of animals occupying a given uniform area; but the propor- tionate composition in both species and individuals is constantly changing, as influenced by neighboring faunae, as well as endemic factors. 16 PROCEEDINGS OF THE ACADEMY OF [Jan., ON THE TERRESTRIAL VERTEBRATES OF PORTIONS OF SOUTHERN NEW MEXICO AND WESTERN TEXAS. BY WITMER STONE AND JAMES A. G. REHN. In the spring and early summer of 1902, Mr. H. L. Viereck and the junior author spent about eleven weeks in, the Sacramento Mountain region of south-central New Mexico. The expedition was undertaken in the interest of the Academy, under the direction of the Board of Cura- tors, and quite extensive collections of plants and insects as well as a fair representation of other groups were secured. The localities visited and represented by specimens in the collections are as follows : Ysleta, El Paso county, Tex. Elevation, 3,664 feet. A village along the Rio Grande a short distance below El Paso. April 2d to 4th was spent here, and collecting carried on in the cultivated strip between the river and sand hills ("bosque"), and also in the latter very un- productive section. El Paso, Tex. Elevation, 3,713-4,000 feet. Collecting was pursued in the foothills of the Franklin Mountains to the north of the city. March 31st and April 4th to 6th were spent in this vicinity. Alamogordo, Otero county, N. M., and vicinity. Elevation, 4,320 to 4,600 feet. This section was quite thoroughly examined, collecting be- ing carried on by both members from April 7 to May 24, and b}^ Mr. Viereck during several days in the earher part of June. Alamogordo is situated in the broad San Augustine plain, two miles from the foothills of the Sacramento range, to which latter almost daily trips were made. The characteristic vegetation of this desert plain is greasewood or creo- sote bush (Larrea) and mesquite (Prosopis). with scattered cactus (Opuntia) and Spanish bayonet (Yucca), the latter becoming more abundant toward the foothills. Dry Canon, Otero county, N. M. Elevation about 4,600-5,000 feet. This canon, being easiest of access from Alamogordo, was visited almost daily. The reptiles, insects and plants of this rather circumscribed region were very thoroughly collected. Alamo Canon, Otero county, N.M. Elevation about 5.000-5.300 feet. This canon is considerably larger than the foregoing and contains a con- stant stream of pure water. It is located about seven miles southeast 1903.]» NATURAL SCIENCES OF PHILADELPHIA. 17 of Alamogordo, and collections were made here April 11, 16, and on several other occasions by Mr. Viereck. Laliiz Canon, Otero county, N. M. Elevation at mouth about 5,000 feet. This canon is about the size of Alamo Cailon, and contains a large stream of discolored alkaline water, locally known as the Laluz river. Laluz Caiion is located seven miles northeast of Alamogordo, and extends a great distance into the range, a portion of it being occu- pied by the railroad which reaches the upper slopes of the mountains. Several days were spent here collecting. Highrolls, Otero county, N. M. Elevation about 7,000 feet. This locality is about half-way up the Sacramento range, and was visited by Mr. Viereck early in June. Cloudcroft. Otero county N. M. Elevation about 9,000 feet. Cloudcroft is located practically on the top of the Sacramentos, and collecting was carried on there by Mr. Viereck during the latter part of May and the early part of June. While the authors are not prepared to enter into a lengthy disserta- tion on the faunal relations of the localities visited, still the general con- clusions drawn from an examination of the material collected may prove worthy of record. Leaving the Texas locahties out of considera- tion, as but little vertebrate material was obtained there, we come first to the immediate vicinity of Alamogordo. This appears to be truly Lower Sonoran (or Middle Sonoran, as has been proposed^), the vege tation, the most characteristic species of which have been men- tioned before, as well as the reptiles, mammals, birds and insects, so far as studied, being characteristic of this zone. This fauna and flora seem to extend quite a distance up the bottoms and arroyos (torrent beds) of the canons, while the slopes of the same and the foothills appear to pos- sess more distinctly Upper Sonoran types, such as the sotol (Dasyli- rion), candle-wood (Fouqueira) and junipers (Juniperus). The pinon (Pinus sp.) does not extend so ar down as the juniper, and makes its appearance only when Lower Sonoran types have totally disappeared from the arroyos. In Dr. C. Hart Merriam's latest faunal map of the United States two tongues of the Lower Sonoran are shown to enter New Mexico from the south, one along the valley of the Rio Grande and the other along the Pecos. From a study of the present material, it would seem that another Lower Sonoran tongue branches off from the ^ The evidence on which this division was made does not appear sufficient. On almost equal grounds we mia;ht divide tlie Carolinian belt into at least two por- tions. 2 18 PROCEEDINGS OF THE ACADEMY OF [Jan., Rio Grande tract and runs up between the Organ and the Sacramento ranges, comprising the San Augustine plain. Highrolls is situated near the lower border of the Transition zone, and possesses few types in common with Alamogordo, but shares quite a number with Cloudcroft. The latter locality is situated in heavy Canadian forests, throughout which are interspersed open glades. The fauna and flora of Cloudcroft are pure Canadian, and the surround- ing timlDerland is one of the most important sources of lumber in the southwest. Considerable material other than that collected by the expedition was examined in the preparation of this paper. In such cases full credit is given to the sources from which these specimens were obtained. We are, however, under especial obligations to Mr. Gerrit S. Miller, Jr., of the United States National MiLseum; Dr. J. A. Allen, of the American Museum of Natural History, New York, and Dr. C. Hart Merriam, Chief of the Biological Survey, United States Department of Agricul- ture, for the loan of specimens of mammals. While concurring in the identifications and statements contained in the following pages, the authors would have it understood that Mr. Rehn is responsible for the portion on mammals and Mr. Stone for the birds and reptiles. MAMMALS. Sciurus fremonti lyclinuclius n. subsp. Type: No. 11,105, Coll. Acad. Nat. Sciences of Philadelphia, c? Forks of Ruidoso, Lincoln county, N. M. August 18, 1898. Collected by C. M. Barber. Related to S. f. ncomexiccntus Allen, but differing in the larger size, more rufous coloration and in the much more elongate nasals. Distr ibid ion. — This form is at present represented only by material from the White ^Mountain region, but probably occurs also in the Cap- itan and Sacramento sections. Color. — Above dull ferruginous, strongest on the median section, the color clearer and most intense on the rump, where the usual faint black- ish vermiculations are almost absent. Sides above the lateral line pale rufescent with an ochraceous tinge; lateral line narrow, not very distinctly defined and with many ochraceous hairs intermingled. Head washed with blackish, the orbital ring clear white; ears externally ferruginous touched with blackish at the apex, which bears a scant pen- cil. Fore limbs clear rufescent above, the lateral fringe of the feet sil- very white. Hind limits dull ferruginous, becoming rufescent on the 1903.] NATURAL SCIKN-CES OF PHILADELPHIA. 19 upper surface of the feet. Tail above centrally rich ferruginous, most intense basally, lateral fringe with the basal portion of the hairs black, the apical section pale ochraceous, the black becoming more extensive toward the apex of the tail and forming an almost clear black bar about 20 mm. in length, the extreme tip being formed by the pale ochraceous tips of the hair. Lower surface of the tail pale grayish ochraceous, the successive black and ochraceous color divisions being marked as above ; toward the apex of the tail a trace of the upper color is noticed suffusing the grayish ochraceous portion. Under surface, including the lips, chin, throat, under surface of the limbs and feet, whitish, pure on the throat, mixed elsewhere with the tint of the pale mouse-colored under fur. Skull. — The skull of S. fremonti hjchnuchus differs from that of S. fremonti neomexicanus in its more elongate form and slender and more attenuate nasals. The skull of hjchnuchus is comparatively narrower than that of neomexicanus, and the zygomatic region appears com- pressed. The nasals average 1.7 mm. longer in lychnuchus than in neomexicanus,, while they really average narrower in width by .2 mm. Measurements. — Flesh dimensions of the type: Total length, 350 mm.; length of tail, 140; length of hind foot, 55. Average of three specimens (flesh measurements) : Total length, 335 (326-350) ; length of tail, 140; length of hind foot, 53.3 (51-55). Average of seven speci- mens of S. fremonti neomexicanus:^ Total length, 321 (308-330) ; length of tail, 133 (127-145) ; length of hind foot, 48.5 (45-52). The type skull of lychnuchus measures as follows: Total length, 50 mm.; basilar length, 42; greatest zygomatic width, 29; length of nasals, 16; width of nasals, 8.2. The average of six skulls of the new form is : Total length, 49 (48-51) ; basilar length, 40.9 (40-42.2) ; greatest zygomatic width, 27.9 (27-29); length of nasals, 15.7 (15.5-16) ; width of nasals, 7.4 (7-8.2). Average of four skulls of neomexicanus:'' Total length, 45.9 (43.5-47.3); length of nasals, 14.3 (13.7-15). The. two skulls of topotypes of neomexicanus examined by us give the follow- ing measurements : Total length, 46.7 (46.5-47); basilar length, 39.7 (39-40.5); greatest zygomatic width, 27.7 (27-28.5); length of nasals, 14; width of nasals, 7.6 (7.5-7.7). Specimens Examined. — Four skins from the type locality, Rio Rui- doso and Mescalero, all in the White Mountains. One of the skins and three additional skulls were from the collection of the American Museum of Natural History. 2 Allen, Bull. Amer. Mus. Nat. Hist., X, p. 291. •' Ibid. 20 PROCEEDINGS OF THE ACADEMY OF [Jail., Remarks. — While no specimens of neomexicanus in summer pelage are available, the great difference in color, aside from the larger size and distinctive cranial characters, lead to the opinion that a very distinct geographic form of the neomexicanus type is represented in this mate- rial. Two topotypes of the latter form from the collection of the Bio- logical Survey were examined in this connection, and also four speci- mens (including the type) of mogollonensis from the American Museum collection. No close relationship exists with the latter form, which is of a more grizzled and smoky coloration, as well as having the skull rather different in character. Eutamias cinereicollis canipes Bailoy. 1902. Eiitn7»ias cinei-eicollis canipes Bailey, Proc. Biol. vSoc. Wash., XV, p. 117, June 2, 1902. (Guadalupe Mts., ElPaso county, Tex.) The material at hand shows that this form is well distributed over the higher portions of the White Mountains, while information from a reli- able source gives it as occurring at Cloudcroft. Material has been examined from the following localities: Forks of Ruidoso, Rio Ruidoso, Mescalero and Hale's Ranch, Ruidoso. Compared with E. cinereicollis, (Allen) this form is readily differenti- ated by the diagnostic characters given by Bailey — the paler and duller ochraceous tints, the white tips to the hair on the upper surface of the tail, and b}^ the grayer feet. One differential character which seems to have been overlooked in the original description is the absence of solid black in the lateral dorsal stripes. In all the specimens (7) examined these are mixed blackish and rusty, while in cinereicollis the centers at least are solid black. Eutamias quadrivittatus (Say). 1823. Sciurus quadrivittatus Say, Long's Expedition to Rocky Mts., II, p. 4.5. (.Vt the exit of tlie Arlvansas river from the foothills, Colorado.) One specimen from Canoncita, above Cleveland, Mora county, N. M., collected by Emerson Atkins, and submitted by Prof. Cockerell, is in- distinguishable from specimens from Boulder, Colo. Citellus spilosoma arens (Bailey). 1902. Spermophilus spilosoma arens Bailev, Proc. Biol. Soc. Wash., XV, p. 118, June 2, 1902. (El Paso, Tex.) Three specimens of this form from the vicinity of Alamogordo, Otero county, N. M., are identical with a series from El Paso and Fort Han- cock, Tex. A young specimen from Mesilla, N. M., is clearly referable to this form. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 21 Citellus marginatus (Bailey). 1902. Spermophilus spilosoma marginatus Bailev, Proc. Biol. Soc. Wash., XV, p. lis, June 2, 1902. (Alpine, Brewster county, Tex.) Two specimens from Pecos, Tex., and vicinity of Alamogordo, Otero county, N. M., agree perfectly with the description of this form. The inference drawn from the occurrence of marginatus in communities with arcns is that they represent distinct species, the coloration being typical as well as the proportions. The length of the hind feet appears to be an excellent diagnostic character. Citellus mexicanus parvidens (Mearns). 1896. Spermophilus tnexicanus parvidens Mearns, Preliminary Diagnoses of New Mammals from the Mexican Border of the United States, p. 1, March 25, 1896. (Fort Clark, Kinney county, Tex.) Three specimens of this form are included in the series examined, all being from Pecos, Tex. Two of the specimens are the property of the Academy, the other belonging to the American Museum. Compared with topotypes the Pecos specimens are seen to be identical, and are easily separated from Brownsville specimens, which latter appear to represent true mexicanus. Citellus grammurus (Say). 1823. S[ciurxts'\ grammurus Say, Long's Exped. Rocky Mts., II, p. 72. (Ar- kansas river, at its exit from the foothills, Colorado.) A specimen from Rio Ruidoso, Lincoln county, appears to represent the typical form. Citellus tridecemlineatus pallidus (Allen)? 1877. [Spermophilus tridecemlineatus'] var. pallidus, Allen, Monogr. N. Amer. Rodentia, p. 872, August, 1877. (Plains of the lower Yellowstone river.) A specimen of the thirteen-lined spermophile from Agua Fria Park, Colfax county, N. M., collected by Emerson Atkins, was submitted by Prof. Cockerell. While by no means true pallidus, still it appears to be closer related to this than to any other form. The condition of the specimen is quite unsatisfactory for conclusive comparison. Cynomys arizonensis Mearns. 1890. Cynomys arizonensis Mearns, Bull. Amer. Mus. Nat. Hist., II, p. 305, February 21, 1890. (Point of mountain, near Wilcox, Ariz.) An adult male of this species from the San Augustine plain (or Tula- rosa Valley), near Alamogordo, Otero county, N. M., appears to be abso- lutely identical with the Arizona form. On comparison the skull is found to be identical with representatives of arizonensis from San Luis Springs, Mexican boundary, determined by Mearns. The length of head and body considerably exceeds the maximum given in the orig- inal description. This individual was from a small colony of about thirty holes. 22 PROCEEDINGS OF THE ACADEMY OF [Jan., Onychomys arcticeps Rhoads. 1898. Onychomys arcticeps Rhoads, Proc. Acad. Nat. Sci. Phila., p. 194. May 3, 1898. (Clapham, Union county, N. M.) A series of eight specimens from the vicinity of Alamogordo, Otero county, N. M., appear to be much closer related to arcticeps than to tor- ridus arenicola Mearns from near El Paso, Tex. The ears are consid- erably blacker than in t. arenicola, and the facial markings are more distinct, while the tail is much shorter. Onychomys ruidosae u. subsp. Type: No. 11,091, Coll. Acad. Nat. Sciences of Philadelphia. 9. Ruidoso, Lincoln county, N. M. September 19, 1898. Collected by C. M. Barber. Apparently related to 0. melanophri/s. but differing in its more red- dish coloration, darker head and ears, and in the narrower inter- ptery gold fossa. Distribution. — Specimens from Hale's Ranch in the Wliite Moun- tain region, have also been examined. Color. — Above dull ferruginous, becoming almost clear rufous on the rump ; head and nape blackish-gray washed with ferruginous ; ears blackish-gray beset with fine silvery-white hairs, which are most abun- dant toward the apex; chin pale gra3"ish-pink. Under surface, includ- ing fore and hind limbs (except posterior surface of hind legs), dull white, the under fur plumbeous. Tail rather thinlv haired, with the upper surface mixed brown and white. Skull. — The skull of ruidosce closely resembles that of melanophrys, but differs in the broader rostrum and nasals, the more truncate occipi- tal region, the narrower interpterygoid fossa and in the stouter and heavier condylar process of the mandible. Measurements. — Flesh dimensions of type: Total length, 156 nun. ; length of tail, 47; length of hind foot, 22. Remarks. — While closely related to the melanophrys group, ruidosce is a very distinct species. Relationship exists with fuliginosus Merriam and torridus Coues, but the former is immediately separated by the slen- derer skull and weaker incisors, beside the narrower rostrum and nasals as well as the smaller auditory bulla. From torridus it is easily differ- entiated by the deeper coloration, the grayer head and the more indefi- nite lateral line, as well as the much greater size. Peromyscus texanus sonoriensis (LeConte). 1853. Hesperomys sonoriensis Le Conte. Proc. Acad. Nat. Sci. Pliila., p. 413. (Santa Cruz, Sonora.) Three specimens from the vicinity of Alamogordo, Otero county. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 23 N. M., appear indistinguishable from specimens of sononensis collected by the Mexican Boundary Survey. Peromyacus truei (Shufeldt). 1885. Hesperomys truei Shufeldt, Proc. U. S. National Museum, VIII, p. 407, September 14, 1885. (Fort Wingate, N. M .) Six specimens from Hale's Ranch, Ruidoso, Lincoln county. X. M., appear identical with a topotype from the collection of the U. S. Na- tional jMuseum. An apparently immature specimen from Las ^^egas, collected b}^ Emerson Atkins, and submitted by Prof. Cockerell. appears to be referable to this form. Peromyscus tornillo Mearns. 1896. Peromyscus iornillo Mearns, Preliminary Diagnoses of Xew Mammals from the Mexican Border of the United States, p. 3, March 25, 1896. (Near El Paso, Tex.) Specimens from Mesilla, Donna Ana county, and Clapham, LTnion county, N. M., are slightly darker than the type of tornillo, though appa- rently referable to the same form. Peromyscus rufinus (Merriam). 1890. Hesperomys leucopvs rufinus Merriam, North American Fauna. No. 3, p. 65, September 11, 1890. (San Francisco Mountain, Ariz.) Six specimens examined from Forks of Ruidoso, Lincoln county, and Mescalero, Otero county, N. M. Sigmodon hispidus berlandieri (Baird). 1855. Sigmodon berlandieri Baird, Proc. Acad. Nat. Sci. Phila., ^'II, p. 333. (Rio Nazas, Coahuila, Mex.) Two specimens examined from Mesilla, Donna Ana county, X. M., and Pecos, Reeves county, T^ex. Neotoma micropus canescens Alleu. 1891. Neotoma micropus canescens Allen, Bull. Amer. Mus. Nat. Hist., Ill, p. 285, June 30, 1891. (North Beaver creek. Panhandle of Oklahoma.) A specimen from Pecos, Reeves county, Tex., differs considerably from a Brownsville, Tex., specimen, and agrees with a specimen from Fort Hancock, Tex., in the much paler coloration. Neotoma mexicana Baird. 1855. Neotoma mexicana Baird, Proc. Acad. Nat Sci. Phila.. ^'11. p. 333. (Mountains near Chihuahua, Mex.) Seven specimens examined from the following localities: Ruidoso, Forks of Ruidoso and Hale's Ranch, Lincoln county, and Mescalero, Otero county, N. M. Cratogeomys castanops (Baird). 1852. Pseudostoma castanops Baird, Rep. Stansbury's Exped. Great Salt Lake, p. 313, June, 1852. (Near Las Animas, Colo.) Three specimens examined from Pecos, Reeves count}', Tex. 24 ■ - PROCEEDINGS OF THE ACADEMY OF [Jan., Thomomys fulvus (Wood house). 1852. Geomys fulrus Woodhouse, Proc. Acad. Nat. Sci. Phila., VI, p. 201. (San Francisco Mountain, Ariz.) Specimens of this form have been examined from Hale's Ranch, Lin- coln county, and Las A'egas, San Miguel county, N. M., the latter col- lected by Emerson Atkins and submitted by Prof. Cockerell. One specimen from the latter locality is uniform plumbeous^ except the feet and mouth, which are white. Thomomys fossor Allen. 1S93. Thomomijs fossor Allen, BuU. Amer. Mus. Nat. Hist., V, p. 51, April 28 , 1893. (Florida, La Plata county, Colo.) One specimen of this species from Agua Fria Park, Colfax county, N. M., collected by Emerson Atkins, w^as submitted by Prof. Cockerell. It agrees with the type except for the more hirsute tail. Dipodomys merriami ambiguus (Merriam). 1890. Dipodomys ambiguus Merriam, North American Fauna, No. 4, p. 42, October 8, 1890. (EI Paso, Tex.) A fine series of sixteen specimens of this form was collected in the vicinity of Alamogordo, Otero county, N. M., and a single specimen was also examined from Pecos, Reeves county, Tex. A considerable range of variation is presented in the series, some specimens appearing quite grayish in comparison to the usual pale buffy coloration. Dipodomys spectabilis -Merriam. 1890. Dipodomys spectabilis Merriam, North American Fauna, No. 4, p. 46. October 8, 1890. (Dos Cabezos, Ariz.) A specimen of this form from vicinity of Alamogordo, Otero county, N. M., probably not quite mature, differs from topotypes from the Biological Survey collection in the slightly smaller size and more gray- ish coloration. Perodipus ordii (Woodhouse). 1853. Dipodomys ordii Woodhouse, Proc. Acad. Nat. Sci. Phila., VI, p. 235. (El Paso, Tex.) A specimen of this species was examined from Pecos, Reeves county, Tex. Perognathus flavus Baird. 1855. PerognatMis flavus Baird, Proc. Acad. Nat. Sci. Phila., VII, p. 332. . (El Paso, Tex.) Six specimens of this form from the vicinity of Alamogordo, Otero county, N. ^L, were examined. 1903.] NATURAL SCIENXES OF PHILADELPHIA. 25 Perognathus penicillatus eremicus (Mearns). 1898. Perognafhus {Chcetodipus) eremicus Mearns, Bull. Amer. Mus Nat Hist., X, p. 300, August 31, 1898. (Fort Hancock, Tex.) Two specimens of this race were collected in the vicinity of Alamo- gordo, Otero county, N. M. Perognathus hispidus paradoxus (Merriam). 1SS9. Perognathus paradoxus Merriam. Xorth American Fauna, No 1 p 24 October '25, 1889. (Banner, Kan.) Two specimens of this form were examined, one from ^Mesilla, Donna Ana county, and the other from Rio Gallinas, Las Vegas, San IMiguel county, N. ]\1., collected by Emerson Atkins and submitted by Prof. Cockerell. Perognathus intermedius Merriam. 1889. Perognathus intermedius Merriam, North American Fauna, \o 1 p 18 October 25, 1889. (Mud Spring, Ariz.) The collection of the Academy contains a male from Mesilla, Donna Ana count}', N. M. Lepus arizonae minor Mearns. 1S96. Lepus arizonce. minor Mearns, Proc. U. S. Nat. Mus., XVIII p 557 June 24, 1896. (El Paso, Tex.) Two specimens from Pecos, Reeves county, Tex., are referable to this form, Lepus texianus griseus Mearns. 1896. Lepus texianus griseus Mearns, Proc. U. S. Nat. Mus., XVHI, p. 562 June 24, 1896. (Fort Hancock, Tex.) An immature specimen from the vicinity of Alamogordo, Otero county, N. M., appears to belong here, but it is not developed enough for positive identification. Putorius (Lutreola) vison (Schreber). 1778. Mustela vison Schreber, Siiugthiere, III, p. 463. (Eastern Canada.) A skin of a mink collected at Las Vegas, San Miguel county, N. M., by Marshall Robbins, was submitted by Prof. Cockerell. In the absence of the skull and strictly comparable material, the exact relationship to the various geographic races cannot be ascertained. Myotis evotis (H. Allen). 1864. Vesperfilio evotis H. Allen, Monogr. Bats N. Amer., p. 48, June, 1864. (Monterey, Cal.) Two specimens of this species have been examined, one from Sapello Canon, Las Vegas range, San Miguel county, N. ]\I., taken at an alti- tude of 10,000 feet; the other from Dry Canon, Sacramento Moun- tains, Otero county, N. M. 26 PROCEEDINGS OF THE ACADEMY OF [Jan., Nyctinomus mexicanus (Saussure). 1860. M[olossus] mexicanus Saussure, Revue et Magasin de Zoologie, 2e s^r., XII, p. 283. (Plateau and higher mountains of Central Mexico, — Cofre de Perote (13,000 feet), Ameca and foot of Popocatepetl (8,500 feet) specified. The name mexicanus is used provisionally for the Nyctinomus of cen- tral and western Texas and southern New Mexico. Further study may show that it is not identical with the central Mexican form, but as far as can be judged by Saussure's measurements, closer relationship exists with mexicanus than with cynoccphalus. The latter form is a smaller animal, the average length of the forearm in five specimens from Tarpon Springs, Fla., being 40.1 mm., while five specimens from New Mexico and Texas average 42.7 mm. True hrasiliensis is, of course, a quite different animal. Specimens have been examined from Mesilla, Donna Ana comity, N. M.; Pecos, Reeves county; Helotes, Bexar county, and San Diego, Duval county, Tex. BIRDS. The birds oljtained number 122 specimens, representing thirty-one species, while ten others are added to the list which were seen but not secured. These latter are marked with an asterisk. Under each species are given Mr. Rehn's field-notes relative to its distribution, abundance, etc. *Querquedula discors (Linn.). "(Jne observed in alkali stream, Laluz Canon, April 18." *Callipepla squamata (Vig.). " Several observed near the mouth of Dry Canon, May 2 and 9." Zenaidura macroura (Linn.). "Observed frequently April 16 to May 19. In the open country it was most frequent about cultivated tracts, and in the canons could almost always be looked for in the vicinity of water." *Catliartes aura (Linn. ) ? " Seen on four occasions, April 18-29. ' ' Dryobates scalaris bairdi (Malh.). "Found frequently in Dry Canon. It was comparatively wild, and showed a decided preference for a peculiar species of cactus." Chordeiles acutipennis texensis (Lawr.). "Found several times in the Larrea belt, and was apparent^ nesting between the grease wood and mesquite bushes. All those seen were flushed from the ground, but careful search failed to reveal the eggs. Mav 2 to Mav 17." 1903.] NATURAL SCIENCES OF PHILADELPHIA. 27 ♦Aeronautes melanoleucus (Baird). "Rather numerous in Dry Canon. April 22 to iMay 20; usually noticed in the morning." Trochilus alexandri Bourc. and Muls. "Several times noticed in Dry Canon, April 17 to May 19, and one female secured. A nest obtained was built upon a small bush growing horizontally from the perpendicular side of a small caiion, about ten feet from the bottom." *Tyrannus verticalis Say? "Several birds apparently of this species were observed May 1 to 2, but were very wild and apparently migrants^ as none were seen later." Myiarchus cineracens (Lawr.). " Observed almost continuously, April 18 to May 17, everywhere in the canons and sometimes in the Larrea as well. Its favorite perch was the top of a spidery cactus or the head of a yucca." Sayornis saya (Bonap.)- " April 22 to May 19. Rather numerous in the canons, near water." ♦Corvus corax sinuatus (Wagl.). "Several observed each morning, April 12 to 17, and again May 15. Seen also at Ysleta, Tex,, April 4." Corvus cryptoleucus Couch. " One shot April 25, a mile and a half north of Alamogordo. It was feeding on the carcass of a cow." Icterus parisorum Bonap. " April 17 to j\Iay 16. Fairly numerous near the arroyos about the mouth of the caiions and in the lower foothills, and very shy." ♦Scolecophagus cyanocephalus (Wagl.). "Several noticed in the town of Alamogordo, April 25 and May 22." Carpodacus mexicanus frontalis (Say). "Rather numerous through April, but not so common in May. Most frequent near the spring in Dry Canon, where they could be seen at almost any time drinkihg from the film of water that trickled down the side of the canon." Astragalinus psaltria (Say). "Seen with the preceding, April 7 to May 17." All the specimens are molting and, contrary to the custom in the Eastern A. tristis, the flight feathers are being renewed at the prenuptial molt. 28 PROCEEDINGS OF THE ACADEMY OF [Jan., Zonotrichia leucophrys (Forst.). "Found both in the canon and on the plain in the Larrea belt, always in small flocks, April 7 to April 29, and several on May 1 ." Some of the males (one-year-old birds?) are in the prenuptial molt and are acciuiring the black and white feathers of the crown. Other highly plvmiaged individuals, probably old lairds, show no signs of molt. Spizella breweri Cass. "One secured April 15, and a numlier seen the next day in the Larrea." Spizella atrigularis (Cab.). One secured in Dry Caiion, May 17. This captiu-e extends the range of the species some distance eastward, the Hachita Mountains being the previous limit of its distribution, so far as I can find in the records. Amphispiza bilineata deserticola Ridgw. "By far the most abundant and characteristic bird of the region, both on the plain and in the caiions. Observed continuously, April 7 to May 21. A nest with three eggs was found May 14, which was possibly a second clutch, as a fledgling was secured some days before." Aimophila ruficeps scottii (Senn.). "Seen several times in the canons, April 11 to May 17." Pipilo fuscus mesoleucus (Baird). " Noticed only in the caiions and on the lower foothills, April 8 to 22." Oreospiza chlorura (Aud.). "Associated with the preceding. April 8 to 29." Calamospiza melanocorys Stejn. A number seen May 1 in the canon. One male secured had nearly finished the molt, including the tertials, but not the rest of the flight feathers. Piranga hepatica Swains. A male obtained in the foothills. May 8. *Petrochelidon lunifrons (Say). "Several observed May 19." Hirnndo erythrogastra Bodd. "Observed several times, April 18, May 20 and 22." *Tacliycineta thalassina (Swains.). "April 7 to May 20; usually seen toward evening." *Lanius ludovicianus excubitorides (Swains.). "Observed at Ysleta, Tex., April 2, and at Alamogordo, May 22." 1903.] NATURAL SCIENCES OF PHILADELPHIA. 29 Dendroica auduboni (Towns.). One secured April IS, Laluz Canon. Helminthophila virginiae (Baird). One shot in Dry Cafion, April 29. Wilsonia pusilla pileolata (Pall.). Two obtained in Dry Canon, April 24 and 29. Mimus polyglottos leucopterus (Vig.). "Abundant about the mouth of the canon, May 8-21." Toxostoma crissalis Henry. "Very retiring in habits and always frequenting the mouth of the caiion, April 7 to May 16." Heleodytes brunneicapillus couesi Mearns. " Noticed in the foothills in small flocks, April 9." The specimen secured seems to be unquestionably this form, although two from Pecos in the Academy's collection approach anthonyi some- what in the broadening of the marks on the back. Salpinctes obsoletus (Say). "Noticed continuously, April 7 to May 21, in the canons only." Tbryomanes bewickii leucogaster (Baird). Two secured in Dry Canon, April 25 and May 6. Catherpes mexicanus conspersus Ridgw. "Found in the deeper parts of the canon, where the junipers begin and the greasewood stops, April 8 to May 17." The wear on the plumage of the back of these birds causes a striking difference in their appearance. First the white borders to the spots are lost and then the spots themselves, leaving the back almost uniform and very different in appearance from the perfect postnuptial plumage. Auriparus fiaviceps (Sundcv.). "Noticed several times in the cailon, April 18 and Mav 9." A nest and set of eggs were secured. Polioptila caerulea obscura Ridgw. "Found almost entirely in the canon, April 24 to May 13." REPTILE3S. Sixteen species of reptiles were obtained, many of them represented by good series of specimens. These are about equally divided between forms which range over the central Lower Sonoran area of Texas and those which are characteristic of the desert region of Arizona to the 30 PROCEEDINGS OF THE ACADEMY OF [Jan., westward, the valleys of southern New Mexico being respectively the western and eastern limit of range of the two groups. Of the species which, according to Cope, range over the first plateau region of Texas (Lower Sonoran) may be mentioned : Sceloporus torquatus poinsetti, Holbrookia texana, ?Sceloponis consobrinus, Salvadora grahaniicc, Anota modesta, Crotalus adamanteus atrox, Phrynosoma cornutum, Crotalus confluentus. The species of the more western desert region are : Cnemidophorus tessellatus, Crotaphytes unslizenii, iCnemidophorus gularis, Uta stanshunjiana, Sceloporus clarkii, Pityophis sayi bcUona. Crotaphytes c. baileyi, Of course, the distribution of some of these species has not been worked out in sufficient detail to warrant positive statements as to their range, while a number of recorded localities are so obviously erroneous that we are forced to attribute them to confusion of labels, an accident which could easily happen in the early days of collecting. However, the geographic relationship of the species contained in the present col- lection would seem to be about as alcove. To add to the value of the following list we have added notes on the species contained in a valuable collection from Pecos, Tex., recently presented to the Academy by Mr. Arthur Erwin Brown. This region being very similar in its faunal relations to the vicinity of Alamogordo, the comparison of the faunae is of much interest. Crotaphytes collaris baileyi (Siejn.). Two specimens secured in Dry Canon, May 10 and 13. Dr. Stejneger has divided the old C. collaris into two very closely related forms, C. baileyi of the western deserts and true C. collaris of the eastern region (Texas). In Prof. Cope's paper on the geographic position of Texas, he describes this species as ranging over what we now know as the Lower Sonoran area of that State, while according to Dr. Merriam C. baileyi in the west- ern deserts is an Upper Sonoran form. It w^ould be interesting to know the exact zonal relationship of the specimens described by Prof. Cope from western Texas and New Mexico, where he states that both forms occur together and exhibit all stages of intergradation,so far as arrange- ment of scales is concerned (Crocodilians, Lizards and Snakes of North America). A series of eleven specimens in Mr. A. E. Brown's collection 1903.] NATURAL SCIENXES OF PHILADELPHIA. 31 from Pecos represents both forms, so that at least C. haileyi must be regarded as a subspecies. Crotaphytes wislizenii B. and G. One obtained on the plain, May 6. This species is regarded by Dr. Merriam as typical of the Lower Sonoran, but also ranging a certain distance into the Upper Sonoran belt. It has been obtained in the vicinity of Pecos, Tex., but not farther east. TJta stansburyana B. and G. A large number seciu-ed in the country about Alamogordo show con- siderable variation in color, but the relationship of the species and its possible races is too involved for present consideration. Holbrookia texana Troschel. This species was common in Dry Caiion. It is a characteristic form of the Lower Sonoran of central Texas, and has been obtained by Prof. Cope as far west as Lake Valley, N. M., which seems to mark the west- ern limit of its range. Mr. Brown's Pecos collection contains several of this species. Sceloporus torquatus poinsettii B. and G. One adult. May 10. This is a common species in central Texas, and has been recorded by Cope from southeastern Arizona. Even if the localities on these speci- mens are correct, it would seem to be decidedly an eastern species. Mr. Brown obtained one from the vicinity of Pecos, Tex. Sceloporus clarkii B. and G. One young individual secured near Alamogordo, May 24. This form is closely related to S. magister, but according to Dr. Stej- neger is restricted to southeastern Arizona (Upper Sonoran), while the latter is characteristic of the Lower Sonoran deserts north to Utah and Nevada. This specimen represents the farthest point east at which the species has been recorded, so far as I am aware, though Prof. Cockerell has sent one to the U. S. National Museum from Las Cruces, N. M. Four specimens are in "Sir. Brown's Pecos collection. Sceloporus consobrinus B. and G. This lizard was common in Dry Canon as well as on the desert. It is apparently the form so identified by Cope, l3ut probably not the one so named by Stejneger in the report on reptiles collected on the San Fran- cisco Moimtain survey. One specimen from HighroUs, Saci'amento Mountains, in the Transition belt, is doubtless referable to the latter. It is darker and more heavily built than those from the desert. The material in the Academy collection has not sufficiently accurate data as 32 PROCEEDINGS OF THE ACADEMY OF [Jan., to locality and altitude to properly work out this difficult group, so that the above identification must be taken as provisional. Found plenti- fully also at Pecos, Tex. Phrynosoma cornutum Harlan. Common on the plain all about Alamogordo; a typical species of the eastern Lower Sonoran region and extending well up the Rio Grande Valley in New Mexico. Phrynosoma douglasii hernandesi Girard. Obtained alDundantly by ]\Ir. Viereck in the Transition and Canadian belts of the Sacramento Mountains. Specimens were also secured for the Academy from Sapello Canon. San Miguel coimtv. N. M.. in August, 1901, by Dr. Henry Skinner. Mr. Viereck brought home a live individual of this species and kept it in captivity for some time. It proved to be an adult female, and gave birth to thirteen young. He gives me the following information rela- tive to the operation : When first noticed four young had been born, a fifth appeared at 10.05 A.M., and after that one was deposited about every five minutes. The birth was effected by the female standing on her hind legs as high from the ground as possible. There first appeared a drop of clear fluid, followed by a bubble-like transparent membrane containing the young, which is then entirely discharged and dropped, the operation taking less than two minutes. If the young is perfect it will soon begin to wriggle, and will clear itself of the membrane in about five minutes from birth. The young are at first narrow and cylindrical, but as soon as they begin to inhale air they become broad and flat like the adult, and darken in color. They measured at birth 16X10 mm. Some were evidently born prematurely and showed no signs of life; they were doubled up and surrounded by a glutinous mass. Anota modesta (Girard). Very common on the plains about Alamogordo. This is another species of the Lower Sonoran of central Texas, and a number are in Mr. Brown's collection from Pecos. Cnemidophorus gularis B. and G. Abundant in the canons, but not found in ine desert regions about Alamogordo, where its place is taken by the following species. To the westward the relative distribution of the two remains the same, C. gularis occupying the Upper Sonoran of Arizona east of Tucson, while C. tes- sellaUis spreads over all the Lower Sonoran desert areas and, according to Dr. Merriam, pushes up some distance into the L'pper Sonoran of the 1903.] NATURAL SCIENCES OF PHILADELPHIA. 33 Death Valley region. The exact relationship and distribution of these two forms and the eastern C. sexlineatus is a very difficult problem, though it would probably be much simplified by adequate material mth exact data, such as that borne by the series before me. Many of the localities quoted by Prof. Cope in his Crocodilians, Lizards and Snakes of N. A. are so exceedingly vague and others so doubtful that they had better be ignored, at least until confirmed. After examining the mate- rial received by him from Texas, it would seem that C. sexlineatus extends over the Austro-riparian region of eastern Texas, departing but little from the normal type until we reach the border of the Lower Sonoran (first plateau) belt. Here we have a form closely resembhng the C. gularis of Dry Canon, which would seem to be the only form in central Texas. C. grahami Cope, of which I have examined speci- mens from the Staked Plains, seems to be most closely related to C. tessellatus; but if the central Texas form is C. gularis, we would have the faunal relation of the representatives of the two species exactly reversed as compared with the vicinity of Alamogordo, since the Staked Plains are Upper Sonoran, and central Texas Lower Sonoran. The Academy has both species from the vicinity of Pecos, but with- out details of distribution; and as both faunal belts occur there in close proximity, it is quite possible that their distribution is as sharply limited as at Alamogordo. Mr. Brown's Pecos collection contains specimens of both tessellatus and gidaris, as well as some identified as sexlineatus. They are accompanied by no data as to relative habitats. Cnemidopliorus tessellatus (Say). Abundant in the desert about Alamogordo. Pityophis sayi bellona B. and G. A number of specimens obtained, which are probably referable to the western desert race. They possess a broad rostral, but have usually a well-indicated black, subcaudal, longitudinal stripe. Salvadora grahamise B. and G. One specimen obtained which belongs to the typical form. Several others were seen. Mr, Brown obtained one from Pecos, Texas. Crotalus adamanteus atrox B. and G. One specimen secured. Crotalus confluentus Say. One specimen which agrees exactly with Prof. Cope's type of C. pul- verulentus from Lake Valley, N. M., but the characters do not seem to have any significance further than indicating individual variation. 3 34 PROCEEDINGS OF THE ACADEMY OF [Jan., The following additional species are contained in Mr. A. E. Brown's collection from Pecos, Tex. Eumeces obsoletus B. and G.. Uta ornata B. and G.. Rhinochilus lecontei B. and G. Heterodon nasicus B. and G.. Zamenis jlageUum Shaw, Zamenis tceniatus ornatus B.andG. Contia episcopa Kenn., Glaucona dulcis B. and G., Tantilla nigriceps Kenn. BATRACHIANS. The only Batrachian obtained was Bufo punctatus B. and G. Two individuals from the spring in Drv' Canon. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 35 MYRIOPODS FROM BEULAH, NEW MEXICO. BY RALPH V. CHAMBERLIN. The chilopods and diplopods noted arid described in this paper com- pose a small collection kindly sent me for determination by Prof. T. D. A. Cockerell, to whom I wish here to express my thanks. All were col- lected at Beulah, N. M., which is at an elevation of about 8,000 feet abovd*sea-level and within the Canadian zone. The forms prove to be largely peculiar, each being easily distinguished from any previously described species. The Geophilus is especially interesting. Lithobius glyptocephalus sp. nov. Diagnosis. — Angles of the 9th, 11th and 13th dorsal plates produced; articles of antennae 20; ocelli 16-18 in three series ; prosternal teeth 6-6, 7-7; spines of first legs beneath 1, 3, 2, of penult 1, 3, 3, 2, of anal 1, 3, 3, 1 ; coxal pores 6, 7, 7, 7-6, 7, 8, 8, transverse; length 20 mm. Description. — Body brown to light-brown, head and posterior dorsal scuta darker, reddish-brown or chestnut; venter paler; legs colored similarly to body, but distal joints, especially the first tarsal, pur- plish ; antennae chestnut proximally , the terminal portion light-brown ; dorsal plates roughened, sparsely pilose, especially caudad; antennae moderately pilose with stiff hairs. Head subcordate, sHghtly wider than long (12: 11.5); a strong fur- row extending dorso-mesad on each side, each furrow dividing into two branches, the anterior of which curves forward to unite with its fellow of the opposite side just back of the frontal suture, the posterior branches uniting similarly in front of caudal margin of head. Ocelli in a narrow elongate patch which is widest behind; 16-18 in number, arranged in three series (1+7, 6, 4-1+5, 5, 5) ; rather large and distinct. Antennae short, consisting of 20 articles of moderate length, the ulti- mate long and pointed at the end. Prosternal teeth 6-6, 7-7, uniform. Angles of the 9th, 11th and 13th dorsal plates produced; plates more or less distinctly depressed along the median longitudinal line ; posterior margin of principal plates, except 7th, sinuate, posterior margin of 7th straight; principal plates strongly marginate laterally and posteriorly. 36 PROCEEDINGS OF THE ACADEMY OF [Jan., Spines of the first legs "•""^■-'■^'•^ ; of the penult '--^i^^', the claw with KJiJi.xi.^>j => 0, n, 1, 3, 2 ' ^ 0, 1, 3. 3, 2' one spine; of the anal ^■-''-f-- 1-/", the claw unarmed; last two coxse armed laterally with a stout spine. Coxal pores 6, 7, 7, 6-6, 7, 8, 8, transversely elongate. Gonopods of female with the claw tripartite, the lateral lobes much reduced, almost minute; basal spines 2-2, stout, clavately thickened distally. Genital appendages of male short and wart-like, pilose. Length of body 20 mm.; width of 10th plate 2.6 mm.; length of antennae 8.1 mm.; of anal legs 8.2 mm. Remarks. — The description is based upon two adult specimens, a male and a female. This species seems to be nearest L. howei Boll., from which, however, it may easily be separated l^y the characters assigned in the diagnosis. Lithobius Beulae sp. nov. Diagnosis. — Angles of none of the dorsal plates produced; head but little wider than long (10.5: 10); articles of antennte 21; ocelli 9-11, in two series; prosternal teeth 2-2; spines of first legs beneath 0. 2, 1, of penult 1, 3, 3, 1, the claw armed with a single spine; posterior coxse unarmed beneath or laterally; coxal pores 2, 3, 3, 3, round; length 8 mm. Description. — Body, head and antenna? brown, legs paler; polished; dorsum uniformly sparsely pilose with short hairs, hairs of venter more scattered; head subglabrous; legs, ]3rosternum and prehensorial feet sparsely pilose; antennse proximally Avith but few hairs or sub- glabrous, distally rather densely pilose; genital and anal segments clothed l^elow with long hairs. Head cordiform, posterior border truncate; a little wider than long (10.5: 10); posterior lateral borders depressed; impressed on posterior portion with two longitudinal diverging sulci, each with a tendency to split into two lines anteriorly. Ocelli in a linear patch, 9-11 in number, arranged in two series (1+3,5-1+4,6). Antenna^ short, articles 21, mostly short, the ultimate long and cylin- drical. Prosternal teeth 2-2, pale, acute. Principal dorsal plates all margined laterally and caudally ; the lateral and posterior borders, especially in the anterior plates, depressed, the dorsum within border strongly arched, bisulcate; posterior borders of lesser plates all straight, of principal ones gently sinuate, none pro- duced, 1903.] NATURAL SCIENCES OF PHILADELPHIA. 37 Spines of the first legs "> Pii-iirlii • of the penult ^^J'jlsjJ'J the claw ^ ° 0,0,0, 2, l' -^ 0.1,3.3,1' armed with a single spine; posterior coxse unarmed beneath or laterally. Coxal pores round, 2, 3, 3, 3. Genital appendages of male small, wart-like. Length of body, 7-8 mm.; width of 10th dorsal plate, .76 mm.; length of antennae, 2.3 mm. Remarks. — The description is based upon two males, both of which have lost the anal legs. The general appearance is not unlike that of L. utahensis Chamb., from the mountains of Utah. Geophilus atopodon sp. nov. Diagnosis. — Frontal plate discrete, last ventral plate narrow; anal pores present, moderately large; pleural pores 10-16 adjacent to ven- tral and dorsal plates ; coxse of prehensorial feet each armed with a stout tooth, the tooth of claw truncate at end; pairs of legs 49-51 (9 ), claw of anal legs long (9 ) ; length, 25 mm. Description. — Rather robust, very gradually attenuated cephalad, more abruptly caudad; color uniformly light-brown, the legs the same, the antennse lighter distally. Dorsal plates smooth, sparsely pilose; legs very sparsely pilose; last seven articles of antennae densely pilose with short hairs, the proximal ones more sparsely provided with long hairs ; prosternum and coxse of prehensorial feet finely punctate. Antennae short, the ultimate joint reduced distally, shorter than the two preceding taken together. Cephalic plate much longer than wide (2.5: 1.9), anterior and pos- terior margins nearly straight, the sides curving, wider in front than behind. Frontal plate discrete, the suture V-shaped in the middle with the point directed backward. Cephahc plate back of suture im- pressed with a strong suture near each lateral margin and with one each side of the median line; a median sulcus extending cephalad in the V- shaped opening of the suture. Prebasal plate concealed. Basal plate as a whole more than twice wnder than long (9:4), exposed portion wider than long in the ratio 9 : 2.25. Claws of prehensorial feet when closed scarcely reaching the end of the first antennal article ; claw armed at base with a moderately stout, truncate tooth; coxa armed with a stout tooth which is somewhat longer than that of the claw; prosternum wider than long (12: 10.5), less than twice the height of the coxa, unarmed. ' Dorsal scuta not sulcate or with a few of the middle ones indis- tinctly bisulcate ; anterior prsescuta short, becoming long or very long in the posterior middle region, the last ones again short. 38 PROCEEDINGS OF THE ACADEMY OF [Jan., First anterior spiracle large, vertically oval, a few following of the same shape, others round, gradually decreasing in size from the first to the last. Anterior ventral plates with a median sidcus, some of the middle ones with a median oval area marked off laterally by semi-lunar depressions. Ventral pores numerous, in more anterior plates arranged in a median depressed area, in first middle plates in a longitudinal median band and in a narrower band cephalad of posterior margin, further back covering nearly the entire surface, although in some more or less absent from a median area while densely covering the plate elsewhere. First pair of legs shorter and much more slender than those succeed- ing, legs increasing in length from the first to the last. Anal legs with a long claw. Pleurae of last segment with 10-16 pores arranged mostly along the margin of the ventral plate or partly covered by the latter, a number (4-5) also adjacent or somewhat overlapped by the dorsal plate. Last ventral plate a little wider than long, about as wide as the one pre- ceding. Pairs of legs 49-51. Length, 25 mm. ; greatest width, 1.2 mm ; length of antennae, 2 mm.; of anal legs, .9 mm. Remarks. — The types are two adult females. They are not very closely related to any other known species. Parajulus ueomexicanus sp. nov. Diagnosis. — First dorsal plate smooth, its sides not at all striate; other segments strongly striate below. Repugnatorial pores small, free from the transverse suture. Last dorsal plate blunt behind, not extending beyond anal valves. Pre-anal scale convexly rounded in front and behind, the curved margins meeting at an acute angle laterally. Anal valves marginate, rugose. Male: Mandibular stipes strongly produced below; the greatly enlarged first pair of legs not at all bent or hamate at end, not tuberculate; copulatory feet ex- posed, the anterior plate set obliquely, very wide, clavately enlarged upward, and produced backward above base posteriorly, the inner border bent inward or backward, indented below top, concealing middle and posterior piece from the front and side. Description. — Slender, smooth and shining, glabrous. General color light-l^rown, transversely banded with dark-brown or blackish, a median dorsal line of same ; a series of suboval light-colored spots along lower part of each side ; toward dorsum within dark band of each seg- ment a short row of light-colored, often confluent, blotches each crossed 1903.] NATURAL SCIENCES OF PHILADELPHIA. -39 by a network of fine dark lines ; in each segment laterad of the dorsal line a short, light-colored transverse Hue; feet pale; first dorsal plate and vertex of head hght-brown, covered with a network of fine dark- brown lines; a dark, blackish band between eyes, with within near middle two light fines diverging cephalad : a fight spot mesad from the base of each antenna ; clypeal region fight-brown ; antemise dark. Head at vertex with a shallow median sulcus. Antennae in length about equaling the width of the body: pilose, the terminal articles densely so ; articles long. OceUi in a large triangular patch, about 50 in number, arranged in 8 series. First dorsal plate smooth, not at all striate; other segments strongly striate beneath, above with numerous longitudinal wavy fines; anterior segments more strongly striate than the posterior. Posterior angle of the last dorsal plate rounded, blunt, but little produced, not extending beyond anal valves. Pre-anal scale convexly rounded in front and behind, indented in the middle behind, the two curved margins meet- ing at an acute angle on each side: mth two long bristles, other- wise glabrous. Anal valves glabrous, marginate; the elevate margins crossed by a series of transverse sulci; valves elsewhere roughened by numerous longitudinal rugae. Repugnatorial pores smaU, not touching transverse sutm-e. Legs rather short, not extending beyond sides of body, sparsely pilose, claw strong; last tarsal joint with a row of stout bristles (pulvilli tar sales) projecting outward on each side (c^). Male: Stipes of mandibles strongly produced below; first pair of legs greatly enlarged, reaching anterior margin of gnathochilariiun, nearly straight, not bent or at aU hamate at end, very sparsely pilose, not tuberculate ; coxae of second pair of legs meeting within and produced forward into a narrow tongue-fike process with a round or button- shaped appendage at end, at base of coxse a median pit which continues forward along process as a shallow furrow; copulatory appendages exposed; anterior plate set somewhat obliquely , very wide, clavately enlarged upward, the outer or posterior border much produced back- ward above base, concealing other pieces from front and side, inner border indented below top, bent inward or backward, plate pilose with numerous long hairs outward from bent border and along a fine below and paraUel with upper margin; middle piece (as also outer one) shorter than the anterior, subtriangular in caudal aspect, acutely pointed, with fine, rather distantly placed teeth along sides which are not strongly bent or roUed, inner basal portion produced forward, a hamate process 40 PROCEEDINGS OF THE ACADEMY OF [Jan., in front of each; outer or posterior piece narrowed above, its top rounded and a little bent, bearing a long bristle, at base also one or several long bristles. Segments 47-57. Length, 36 mm.; width, 1.9 mm. Remarks. — The specimens studied include three adults, two males and a female, and several immature individuals. The measurements given are from the largest individual, a male. Polydesmus sp. Several young individuals of a species of this genus are in the col- lection. They are of the stage possessing only twelve segments, and thus too young for determination. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 41 LIST OF THE POLYCYSTID GREGARINES OF THE UNITED STATES. BY HOWARD CRAWLEY. I. The polycystid gregarines of the United States have as yet been but very Httle studied, and the Hst here given, which I beheve to be com- plete, contains but twenty-eight species. Of these, six are also found in Europe and were first described from there. It is to Prof. Leidy that we owe the better part of our knowledge of the American forms. Ten species were described by him in accounts pubhshed in the Proceedings of the Academy and in the Transactions of the American Philosophical Society from 1848-1889. In addition to these, four more species are described and figured by Leidy in manuscript notes now in posses- sion of the Academy, access to which I owe to the kindness of the Secretary, Dr. Nolan. Of the remaining eight forms, one was de- scribed by Porter, and seven were found by me in animals collected at Wyncote, Pa., and Raleigh, N. C, during the past year. The four forms described in Leidy's manuscript, and the one by Por- ter, were not named. I have accordingly created new species for them which, with the seven that I discovered, makes twelve new species. Although the method of naming a gregarine after its host has been ob- jected to by Schneider, it appears to me the most advisable, and it has accordingly been followed. Considerable difficulty was experienced, however, in determining to wliich genera these new species belong. The generic characters of gregarines are mostly those of the cysts and spores, and I was particularly unfortunate in failing to obtain these stages. In several cases it has been possible to judge, even mthout these criteria, what the correct genus is, and such forms have been placed where they probably belong without question. Where, how- ever, there was considerable doubt, a question mark has been placed before the generic name. II. The following is a list of the Arthropoda found to contain gregarines, the first column giving the number of individuals examined, the second the number parasitized : 42 PROCEEDINGS OF THE ACADEMY OF [Jan., Julus and Parajulus 165 158 Polydesmus virginiensis ._.. 32 26 Fontaria sp 2 2 Lysiopetalum lactarium 16 6 Spirobolus sp --— 1 1 Lithohius forcipatus -— 13 4 Scolopocryptops sexspinosus 6 3 Scutigera forceps — 4 3 Geophilus sp 8 Elater sp 8 1 Cucujid larva..- 5 2 Disccelus ovalis — 2 2 Harpalus caliginosus 1 ■ 1 Dytiscus sp., larva 1 1 It is headed with Julus and Parajulus. These two genera of Diplo- pods are distinguished by characters of the mouth-parts and anterior feet, but they are much alike and the smaller species of Parajulus are but little larger than those of Julus. No attempt was made to distin- guish between these two genera and much less between the various species, but such an attempt w^ould have been of no especial utility, even if successful. They are to be found together, beneath logs, stones and leaves, and, as the list shows, they are almost invariably para- sitized, the gregarine being Stenophora juli. It is possible to obtain these Diplopods at any season of the year, and so far as my observa- tions go, the gregarines have no seasonal cycle, but all stages are met with at any time. Usually a given host contains a moderate number of gregarines, from twenty to forty, these ranging from the largest adults to the smallest intracellular stages. Frequently, however, only a very few parasites are found in a host, and in such cases they will be for the most part quite small. In none of the millipedes ex- amined have I come across the cysts, nor, to my knowledge, the free spores. In consideration of their gregarious habit, and their diet of rotten wood and vegetable fibre, it is easy to see why Julus and Parajulus are so persistently parasitized. It is also worthy of note that other animals having the same mode of life and generally found in the same places do not appear ever to contain gregarines. A case in point is that of the Isoi^oda, Oniscus and Porcellio. These are almost always found with Julus and Parajulus, and they doubtless frequently swallow the spores of Stenophora. The isopod intestine is, however, lined with chitin, and 1903.] NATURAL SCIENCES OF PHILADELPHIA. 43 the immunity is probably due to the inabiUty of a gTCgarine sporozoite to get into the cells. The 32 specimens of Polydesmus examined were some from Wyncote and some from Raleigh, N. C. Polydesmus is infected by two greg- arines, Gregarina polydesmivirginiensis and Amphoroides jontarice. Of these, both were present in 14 hosts, G. polydesmivirgimensis alone in five and A. fontarice alone in seven. The Diplopods from one region were about as frequently parasitized as from the other, but G. polydes- mivirginiensis was present oftener and in larger numbers in the animals from Wyncote than in those from Raleigh, while the reverse was true in the case of A. fontarice. It is for this reason that the latter, although a frequent parasite of Polydesmus. has been named for Fontaria. Where two gregarines parasitize a given host, it is generally found that one occurs far more frequently than the other. The more frequent para- site is spoken of as the primary, the other as the secondary. Usually- also, the primary parasite of one host is the secondary of another. In Pennsylvania, where Polydesmus is common and Fontaria does not occur, the former, as stated above, is more often parasitized by G. poly- desmivirginiensis than by A. fontarice. In North Carolina, where Fon- taria is common, the more usual parasite of Polydesmus is A. fontarice. This gregarine, furthermore, was present in great numbers in the two individuals of Fontaria I was able to examine. It may therefore be concluded that G. polydesmivirginiensis is the primary of Polydesmus, while A. fontarice is the primary of Fontaria. In Polydesmus, G. polydesmivirginiensis was usually present in mod- erate numbers, from 20 to 50, although occasionally there were only a very few. On the other hand, A. fontarice ran to extremes. Some millipedes would show only two or three gregarines, while in other cases they were present by the hundreds. There were only a few, however, in the animals taken at Wyncote. Of sixteen specimens of Lysiopetalum lactarium. six were parasitized. Two gregarines occur, one a new species which I have named Gregarina calverti, the other apparently StenopJwra juli. Both gregarines were present in two hosts, G. calverti alone in three and S. jidi alone in one. Thus the former species was present five times and the latter three. G. calverti is therefore to be regarded as the primary parasite of Lysio- petalum. This miUipede is found under stones in much dryer places than Julus, Parajulus or Polydesmus, and it is somewhat solitary in habit. It is thus easy to see why only a comparatively small percentage of them is parasitized. G. calverti occurred in moderate numbers in all 44 PROCEEDINGS OF THE ACADEMY OF [Jan. cases, whereas S. juli was present in hundreds in one case and in only small numbers in the other two. Echinomera hispida, the primary parasite of Lithohius forcipatus, was found only four times in thirteen hosts collected at Wyncote and Ral- eigh. In two of these cases, moreover, but a single gregarine was found. At Cambridge, Mass., where Lithobius is a much commoner animal, a larger percentage are parasitized and the gregarines are present in larger numbers. Actinocephalus dujardini, the secondary parasite of Lithohius, is very rare. Scolopocryptops is parasitized by Hoplorhynchiis actinotus . The greg- arines occur in small numbers, from ten to twelve, in fifty per cent, of the Chilopods. Of the five specimens of Scutigera forceps, four contained gregarines The species is Trichorhynchus pidcher; and never but a very few, from three to ten, are present. It is not necessary to go over the balance of the list in detail. Atten- tion may, however, be called to the fact that eight individuals of Geophi- lus were opened in vain. III. Gregarina blattarum Sieb. Gregarina blattarum Siebold (1839), p. 57, Taf. 3, figs. 57-61. Gregarina blattarum Frantzius (1S48), p. 190, Taf. 7, fig. iii. Gregarina blattarum Stein (1848), p. 182, Taf. 9, figs. 38, 39. Gregarina blattce-orientalis Leidy (1853), p. 239, PI. 11, figs. 39-41. Clepsidrina blattarum Schneider (1875), p. 580, PL 17, figs. 11, 12. Clepsidrina blattarum. Btitschli (1881), p. 384, Taf. 20, 21 Clepsidrina blattarum Wolters (1891), p. 115, Taf. 7. Clepsidrina blattarum Marshall (1893), p. 25, Taf. 20, 21. Common in Periplaneta orientalis, Periplaneta americana and Ectohia (Blatta) germanica. A few specimens of Ischnoptera pennsylvanicus , the field cockroach, were examined, but none contained gregarines. ? Gregarina termitis Leidy. Gregarina termitis Leidy (1881), p. 441, fig. 27. Gregarina termitis Porter (1897), p. 65, PI. 6, figs. 73-76. Leidy says : "A small gregarine was once noticed in our Termite, and is represented in fig. 27. The body is ovoid, with the narrower end pos- terior. The head [protomerite] spheroid, compressed from above down- ward. Length .06 mm. ; head .018 mm, long, .03 broad; body .036 mm . broad." Porter adds the information that the gregarines are very common in some specimens of termites, occurring mostly in the anterior part of the small intestine, and that the nucleus contains one karyosome. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 45 I have opened perhaps a dozen termites in a search for this parasite, but have not yet encountered it. Gregarina achetaeabbreviatae Leidy. PI. HI, figs. 34-35. Gregarina achetceabbreviatce Leidy (18.53), p. 238, PI. 11, figs 33, 34. Epimerite: Simple, spherical. Protomerite: Somewhat pentagonal to hemispherical. Separated from the deutomerite by a sharp constriction. Shape modified in the associations. Deutomerite: Cylindrical, tapering slightly. Terminating bluntly. Epicyte: Thick. Sarcocyte: ?. Entocyte: Of equal density in both protomerite and deutomerite. Nucleus: Large, spherical with several karyosomes. Dimensions: Maximum length 600 microns. Host: Acheta abbreviata. Infection extensive. Locality: Beach Haven, N. J. ? Gregarina scarabeirelicti Leiily. Gregarina scarabeirelicti Leidy (1851), p. 208, 287. Gregarina scarabeirelicti Leidy (1856), p. 47. I have not seen this species. ■? Gregarina meloIontbaebrunneaB Leidy. Gregarina melolonthoebrunneoe Leidy (1856), p. 47. I have not seen this species. Gregarina passalicornuti Leidy. PI. II, fig. 24. Gregarina passalicornuti Leidy (1853), p. 238, PL 11, figs. 30-31. Epimerite: ?. Protomerite: Hemispherical in the primites ; considerably compressed in the satellites. Deutomerite: Cyhndrical; sometimes slightly narrower about the middle. Epicyte: Thick. Sarcocyte: Thick. Entocyte: Opaque. Nucleus: Moderate. Spherical. Dimensions: Average length of pairs 350-400 microns. Host: Passalus cornutus. ■; Gregarina polydesmivirginiensis Leidy. PI. II, fig. 25. Gregarina 'polydesmivirginiensis Leidy (1853), p. 238, PI. 10, figs. 23-29. Epimerite: Not seen. Protomerite: Always small; sometimes apparently undeveloped in 46 PROCEEDINGS OF THE ACADEMY OF [Jan.. the younger animals. Never any constriction between it and the deu- tomerite in relaxed animals . Deutomerite: Long. Anterior half usually much broader than pos- terior half. Usually considerably curved. Epicyte: Very distinct. H-2 microns thick. Sarcocyte: Distinct. EntocTjte: Usually, but not always, considerably denser in the ante- rior part of the deutomerite than in the posterior part . Nucleus: Always with one karyosome . Dimensions: 400 microns long. Movements: This is a very active gregarine, and displays continually both changes of shape and progression. The contractions are generally confined to the anterior part of the deutomerite. As a rule, the smaller animals are less active than the larger. Host: Polydesmus virginiensis. Infection is common, the parasites being present in moderate numbers. Locality: Wyncote, Pa., and Raleigh, N. C. 1 Gregarina elaterae sp. n. PI. I, tig. ii. Epimerite: Spherical, with a granular contents. Usual size and appearance shown by fig. 11. In several cases, however, apparently on account of the absorption of water, it had increased enormotisly in size, so that its diameter exceeded the length of the gregarine. Protomerite: Elliptical in outline, with the greater axis transverse. Posterior limit straight. Separated from the deutomerite by a sharp constriction. Deutomerite: Oval to subspherical. Epicyte: Very thin. Sarcocyte: ?. Entocyte: Granular in all three segments . Sparsely filled with large and verv^ distinct granules, so that the appearance presented was rather that of a ciliate or a heliozoan packed with plant spores than a greg- arine. Nucleus: Not seen. Dimensions: Maximum length 62 microns. Movements: The animals were very sluggish, the only movements seen being an occasional lateral displacement of the protomerite, and, in one case, progression of perhaps one-half the animal's length. Host: The larva of a large species of Elaier. Infection only occa- sional, and never but a few gregarines prpsent . Locality: Wyncote. Pa. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 47 ? Gregarina discaeli sp. n. PL I, figs. 7-10. Epimerite: ?. Protomerite: Helmet-shaped to pentagonal. Separated, from the deutomerite by a shallow constriction in the adults, by a deep constric- tion in the young. Deutomerite: Greatty elongated, cylindrical to slightly conical, termi- nating bluntly. Epicyte: Very thin. Longitudinal striations not evident. Sarcocyte: Apparently wanting. Endocyte: Densely granular and opaque in the deutomerite; nearly transparent in the protomerite. Anterior part of protomerite free from granules. Nucleus: Not readily seen, but apparently possessed of a number of karyosomes disposed in a cluster. Dimensions: Maximum length 1200 microns. In the young the ratio of length to breadth is about eight to one, in the adults about fifteen to one. Movements: These gregarines possess the flexibility of an eel. They frequently bent themselves into crescents, circles and short, close spi- rals, this last phenomenon demonstrating the spirality of the myocyte. Fig. 10 shows the longitudinal axis of one contorted animal observed, A being the anterior and P the posterior end. Host: Disccelus ovalis. Of the two beetles examined, one contained from 50-75 parasites; the other not nearly so many. Locality: Wyncote, Pa. No cases of association were seen, but for the most part these greg- arines were disposed in dense bunches with their posterior ends glued together. There was one such mass seen which presented the appear- ance of a bunch of celery, the animals being all closely apposed from their posterior ends nearly as far forward as their protomerites. 1 Gregarina xylopini sp. n.i PL III, figs. 20, 30. The two gregarines shown in figs. 29 and 30 are stated by Leidy to be parasites of the beetle Xylopinus saperdoides. Of the six beetles exam- ined, five contained gregarines of the form shown in fig. 29, one of the form shown in fig. 30. These two forms are so dissimilar that it appears better, at present, to give only the figures, reserving the description imtil additional information is at hand. ? Gregarina boletophagi sp. n.- PL IT, figs. 20-2^. Epimerite: ?. Protomerite: Large, variable in shape. Separated from deutomerite by a sharp constriction. ' From Leidy's MSS. '' From Leidy's MSS. 48 PROCEEDINGS OF THE ACADEMY OF [Jan., Deutoynerite: Cylindrical, with a protrusion at the posterior end, to conical with a pointed end. Epicyte: Thick. Sarcocyte: ?. Entocyte: Dense, uniform in both protomerite and deutomerite. Nucleus: Moderate ; oval to spherical, with one karyosome. Dimensions: 320 microns long. Host: Boletophagus cornutus. Locality: Swarthmore, Pa. ? Gregarina calverti' sp- n. PI. II, figs. 19-21. Epimej'ite: Not seen. Protomerite: Greatly compressed in the large animals (figs. 19, 20). Anterior boundary usually straight or slightly convex, but frequently concave, giving the protomerite the form of a shallow cup. Approxi- mately reniform in the young animals (fig. 21). Constriction always deep and sharp. Deutomerite: Displays a shoulder at the anterior end, being here widest. Narrows gradually backward, more regularly in the younger, less regularly in the older animals. Posterior end always blunt. Epicyte: Thicker than the sarcocyte. Longitudinal sculpturing may or may not be present. It was often possible to see that the proto- merite was longitudinally striated, but the cause for this appearance seemed to he deeper than the epicyte. Sarcocyte: Thinner than the epicyte. Myocyte: Easily seen in the living animals. Entocyte: In large animals black in transmitted light in the deuto- merite and nearly so in the protomerite. In most cases, however, the granules do not fill the entire entocyte, but thin out rapidly toward the posterior end, so that in the last fourth or fifth of the entocyte they con- stitute only a narrow, central strip. This condition is shown in fig. 21, from a young animal, and is held true for nearly all the gregarines seen. Nucleus: Not large, spherical, with one spherical karyosome. Movements: These gregarines, although quite flexible, are rather sluggish, and usually lie motionless. Progression, when observed, was generally accompanied by lateral movements of the protomerite. As a rule, the animals were slightly curved, a phenomenon frequently dis- played by elongated gregarines. Diynensions: Maximum length 1,500 microns. Host: Lysiopetalum lactarium. Locality: Wyncote, Pa. ''To Philip P. Calvert. 1903.] NATURAL SCIENCES OF PHILADELPHIA, 49 ■? Gregarina harpali sp. n. PI. I, figs. 1-4. Epimerite: Not seen. Protomerite: In the adults (fig. 1) hemispherical and narrower than the deutomerite, from which it is separated by a sharp, but not exten- sive, constriction. In the young (figs. 2 and 3) knob-shaped, wider than the deutomerite and separated from that part by a shallow con- striction. Deutomerite: Conical in the adults to C3dindrical in the young. The posterior end is normally blunt and rounded. These gregarines, how- ever, are quite polymorphic and readily change shape. Epicyte: Relatively very thin in the adults. INIarked with very fine longitudinal ridges, which are not visible in the living animals and require sections for their demonstration. Sarcocyte: Not visible in adults. In the young (fig. 3) present in the protomerite and constituting the thick septum. Apparently wanting in the rest of the deutomerite. In the front part of the protomerite a little papilla is present, apparently composed of a differentiation of the sarcocyte. This papilla is much more evident in some specimens than in others, and sometimes takes the form of a minute knob. It was never seen in adult animals. Entocyte: Black in transmitted light in the adults, semi-transparent in the young, but in both as dense in the protomerite as in the deuto- merite. In the young, as shown in fig. 3, there is a crescentic area almost free from granules. In stained and sectioned animals (fig. 4) this crescentic area is seen to consist of very dense cytoplasm, with a considerable affinity for stains. This condition is wholly lost in the adults. Nucleus: In the living animals, wholly invisible in the adults and not satisfactorily to be made out in the young. Sectioned material shows that it is large, spherical, and possesses several karyosomes. These are variable in size and are scattered singty throughout the entire extent of the nucleus. The}^ display the usual phenomenon of vacuoles. Movements: The movements displayed by these gregarines were leisurely muscular contractions and slow progression. There was also a disposition to rumple the edges of the body, so that the epicyte of the deutomerite presented a series of scallops. These scallops underwent slow changes, the movement being doubtless the usual peristalsis, although very much slower than in the case of most gregarines. Pro- gression was in either more or less of a straight line, or in curves of short radius. In the latter case, the animals held their bodies so Ijent that the longitudinal axis formed a curve. In all cases progression was 4 50 PROCEEDINGS OF THE ACADEMY OF [Jan., extremely slow. A number of animals showed no movement whatever and some of these held the scalloped outhne without change. Dimensions: Length 225-700 microns. _ Host- Harpalus caliqinosus. The gregarines were present m the intestine of the one beetle examined in hundreds, and while no associ- ations were seen, there was a very marked tendency for them to be lying massed closely together. Locality: Wyncote, Pa. mirmocystis ovalis sp. u. PL I, Ags- 5-6. Evimerite: Usuallv elliptical in outhne, with the longer axis trans- verse, but frequently considerably flattened down. Wall with double contour. Contents hyaline. Protomerite: Hemispherical. Deutomerite: Cylindrical to oval, but more usually the latter. Always terminating bluntly. _ Epicyte: Thin over the anterior part of the protomerite; elsewhere thick. Longitudinal striations not seen. Sarcocyte: ?. . , , i i? ; Entocyte: Dark brown in the larger individuals, and of approxi- mately the same density in both protomerite and deutomerite. The anterior third of the protomerite is usually, but not always, free from granules. Nucleus: Not seen. Dmensions; Greatest length 70 microns. ■ , ,. Host- The larva of a beetle doubtfully identified as belonging to the Cucuiidffi The gregarines occurred sparingly, ten or twelve being the greatest number lound in any one host. About half the insects exam- ined w^ere parasitized. T Euspora lucani sp- n.^ PI. HI, fig. 38. lZomt%e:'^m^\\, compressed from before backward. Separated from the deutomerite by a deep constriction. Deutomerite: Cyhndrical; sometimes narrower m the middle. Epicyte: Thick. Seircocyte: ?. Entocyte: ?. d"*«,'o,L- Of one association; primite 320X128 microns, satellite 360 X 1 OS microns . *FromLeidy'sMSS. 1903.] NATURAL SCIEXCES OF PHILADELPHIA. 51 Host: Lucaniis dama . Locality: Swarthmore, Pa. The gregarines occurred either singly or associated in pairs. Stenophora juli Frantzius. Gregarinajuli Frantzius (1848), p. 194, Taf. 7 fig x 1-2 Gregranria iaruateLeidy (1848-49), p. 232. ' *' ' Gregarinajuli-viarginatil.eidy{li5i),Y>.'2Z7,V\ 10 figs l-'^O Gregranm;«Zf-jOimoung upon a nest instead of carrying upon her body, is shared by our loca Dolomedes urinator Hentz, but is different from the usage of the other Lycosid genera. These web-nests are generally found on he top ot plants along the sides of wooded streams, and have a decided resem- blance to the nests of Lepidopterous larva?. Tegenaria derhami Scop. Individuals observed, males: No. 23, captured May 2 escaped May 22; No 74 captured May 13, died (from starvation) June 5; No. 80, captured May 14, kd ed May 27; No' 178, captured June 5, died July 24 No. 218, captured W 9 died abmxt September 1; No 221, captured June 19, died July b. No. -o-, cap tured June 26, killed (by ^ No. 25) the same day. Ind^riduals observed, females: No. 22, captured May 2, died Ma> 21 No. 5, captured May 2, died August 4; No. 75, captured May 13, died August 19 nT. 79, captured May 14, died September 1; No 84, capture^ May 15 kill d May 25; No. 117, captured May 24, died about September 10 No. 118 /^^P^^;^^^ May 24 escaped June 5; No. 219, captured June 19, died in December, No. 220, captured June 19, died the first week of September. Moulting. -Tl^is was seen only once; the moulted skin was split in the horizontal plane as in other spiders. Mating.-The mode of observation was to put a male upon the web of the female, and the male was in most cases left with her and not removed, since removal would necessitate a breaking of the web The male is of about the same size and strength as the female, and the two generally live peacefuUy together. Sometimes the male is the more aggressive and secures all the food placed in the cage, but more generally the female lords it a little over the male. The following are my main observations on the approach of the male and the mode of copulation: , o .i (1) On to the web of 9 No. 22 I placed d^ No. 23 on May 3, at 2.41 P M They touched each other a few times, then remained motion- less facing each other from 2.49 to 3.02. Then she moved toward him and followed him slowly, she finally touched him and he moved away Then both remained immovable until 4.00, when I removed him On May 3 I put him in again at 9.15 P.M.; he touched her, 1903.] NATURAL SCIENCES OF PHILADELPHIA. 93 she instantly turned and faced him, their fore-legs almost in contact. They remained in this position until 9.29, he tapping the web a few times with his palpi, while she shook the web vigorously with her left fore-leg for a couple of seconds. At 9.30 he rubbed his palpi vigor- ously with his fore-legs, then approached and copulated, this act lasting only 5 seconds. Their exact position in the act was not determined. I could only observe that he approached head on with one palpus outstretched, they mutually grasping each other with their first pair of legs, both in their usual position on the upper surface of the web. She then moved away. At 9.35 he walked over her without copulating. At 10.25 she approached and tapped him sev- eral times with her left fore-leg, and again at 11.02. I watched them continuously up to 12.27, then left him with her. Up to May 21, when she died (he securing most of the food given them), I watched them many times, saw him on some occasions approaching her with shaking palpi, but observed no further copulation. (2) To ? No. 25, after she had made 3 cocoons, I introduced d^ No. 225; she chased and killed him. (3) 9 No. 79 had made 3 cocoons, when c? No. 178 w^as introduced on June 19, at 9.15 P.M. He moved toward her very slowly, a step of a centimeter each time, slowly and cautiously tapping the web with his palpi; he took about 9 minutes to cross a distance of 2 inches, and then when he had almost reached her, she rushed at him and drove him away. At 9.35 he again approached cautiously, and at 9.50 rushed at her, but she repelled him. At 9.57 he advanced again, at 10.10 making a cpiick rush at her and almost succeeding in inserting an extended palpus in her epigynum, but she escaped from him. No copulation was seen up to 11.25. (4) 9 No. 117 made her first cocoon on June 10. On June 19, at 9.40, c? No, 221 was placed on her web. At first she chased him about, then both became quiet. At 10.01 he quickly ran at her, and copu- lated with his right palpus from 10.02 to 10.03, for ten minutes after which she chased him about the cage. The copulatory position was the same as that of ? No. 75 X cJ" No. 218. (5) ? No. 75, after making 2 cocoons, mated with d"' No. 218 on June 19. He was introduced at 9.08 P.M., and shortly after she chased him. At 10.00 he made a sudden move toward her and succeeded in copulation; he inserted his right palpus into her epigynum for 1 minute, his left for 15 seconds, his right again for a minute and a half. Position: the male braced himself firmly on all legs upon the web, and reached his palpus straight out before him and pressed it into the 94 PROCEEDINGS OF THE ACADEMY OF [Jan., epigyniim with considerable energy; the female faced him also on the upper surface of the web (just within the funnel), lying partly on one side with her legs drawn up, but not closely, to the sides of her body. During the act the male pushed the female backward by his force. When the palpus was inserted there was seen a large dilated sac evaginated from the palpal organ. The female then ran off, and the male proceeded to the process of sperm-induction. In these copulations the male accordingly first approaches slowly, the female also sometimes making advances on her part, and con- cludes with a rush at the female, and if he is quicker than she he suc- ceeds in inserting his palpus. The fact that females copulate after cocooning makes it appear probable that a copulation may precede the making of each cocoon. Sperm-induction. — 6^ No. 218 was watched during this act, which occurred immediately after his copulation with $ No. 75 (concluded at 10.03 P.M.). He first spun across an area of nearly a square inch on the upper surface of the web, then limited himself to a small area just at the entrance of the tunnel of the web. Here, by very rapid brushing of the spinnerets from side to side, he constructed a deli- cate silk sheeting placed at an acute angle to the surface of the nearly horizontal web, the posterior and lateral edges of this sheeting attached to the web, but its anterior edge unattached and elevated like an arch. From side to side this sheeting was not quite so long as the length of his body, and half this distance from Ijefore backw^ard. He then stood over it, so that the ventral surface of his abdomen almost touched the superior surface of the sheeting, his cephalothorax above the anterior (free) edge of it, and his spinnerets at its posterior edge. At 10.21^ P.M. he moved his abdomen slightly forward, discharged from his genital aperture a minute drop of sperm upon the superior surface of the sheeting at its free edge, and from that moment until 10.28 he was engaged in taking this sperm into his palpal organs. This he did by pressing the ventral (posterior) surface of each palpal organ on the under surface of the sheeting, just at the point where the drop of sperm had been placed on its upper surface; one palpus he held thus, with occasional shaking, for 10 to 15 seconds, then the other, until the whole of the drop was inducted; and each palpus when not so busied he held in front of his cephalothorax and shook it in the air, as if to force the semen deeper into the organ. From 10.27 until 10.34, just after the completion of this process, he remained in the same posi- tion, but quietly and with no further discharge of semen upon the sheeting. Then he left the sheeting, worked his palpi a short while 1903.] NATURAL SCIENCES OF PHILADELPHIA. 95 with his chelicera, then remained quiet until 11.10 when I ceased the observations. The spirit is always willing to watch longer, but con- tinuous observation with a hand lens, and the necessity of keeping quiet in a cramped position for fear of startling the object, becomes trying to the flesh. A curious act which I observed d^ No. 23 to do I cannot explain, but it may possibly have some connection with the mating. On May 3 he was upon the web of 9 No. 22. At 9.50 P.M., after his copu- lation with her (described above), he commenced to tear a hole in the horizontal portion of her web, tearing with his palpi and putting the torn shreds in his jaws. He thus made an oblong rectangular hole of about the length of his body, and as he continued to tear the portion just in front of him, he swung his spinnerets over the posterior part of the hole, and so spun it over with very fine transverse lines. This act lasted until 10.04, and only about one-quarter of the hole in the web had been mended over. Though I watched him closely up to 12.27 A.M., he did not return to this hole. Cocooning. — This was observed several times, and was as follows : 9 No. 79 was observed at 8.00 A.M., June 19, spinning over a cir- cular area (of greater diameter than her own length) on the vertical glass wall just above the edge of the web. She ceased until 8.09, owing to a jar to the cage, then commenced again. Very regularly and actively, without interruption, she spun up to 8.23, placing her feet at the periphery of the silken disk ("base" of the cocoon), and revolving her body over it, turning sometimes from right to left and sometimes in the reverse direction, all the while beating the tips of her palpi upon the disk. During the earlier part of the process the outstretched spinnerets were brushed from side to side ; but in the latter part she lifted the abdomen high after each application, thus pulling out long threads which she fastened close to their first point of attach- ment, so that the inmost (most lately made) part of the disk was made of curled loops, and so was of very soft consistency. Thus the base of the cocoon, somewhat thickest in the centre, was formed. At 8.23 she stood quietly over this base, then oviposited upon it, the ovi- position lasting a little more than one minute. From her genital aperture exuded a large clear drop of viscid fluid and fell upon the centre of the base, its upper surface still connected with her genital aperture, and at the same time the yellowish ova fell into and were enveloped by this drop. Then she immediately begaif spinning the cover of the cocoon over the egg mass. From 8.24| to 8.35 she spun a loose covering of silken loops by elevating the spinnerets after each 96 PROCEEDINGS OF THE ACADEMY OF [Jan., application (just as in the making of the base), and from 8.35 to 8.44 she spun a denser covering by brushing the spinnerets from side to side. Finally she alternated the spinning upon the cocoon itself, with spinning attachment lines from it to web. ? No. 25 made her first cocoon on the horizontal surface of her web, apparently not spinning a special base. ? No. 75 made her second cocoon upon the surface of the web, and after spinning a discoidal base oviposited upon it at 7.41 A.M.; she occupied the time from 7.42 to 8.00 in spinning the cover. ? No. 220 spun her cocoon in the same manner as did No. 79. One female, after making her cocoon at the angle of the wall and floor of the cage, spun over it several thin, concentric, arched lamellae of silk, evidently a protection. In a few cases the female attaches foreign particles, as the remains of insects, to the surface of the cocoons, but this was not done with the majority of the cocoons. The cocoons, even those of the same spider, are placed in most diverse positions, some horizontally and some vertically, some upon the web and some away from it. The number of cocoons made and the dates of construction were as follows: (1) $No. 25: May 31 (hatched June 9); May 29; June 10 or 11 (hatched July 5) ; June 20 (hatched July 16) ; July 2 (did not hatch) ; July 9 (hatched July 31) ; July 21 (not all the eggs hatched) ; August 1 (hatched). A total of 8 cocoons made by an unusually small female. (2) 9 No. 75 : May 23 (hatched June 14) ; June 7 (destroyed by the mother); July 5 (hatched July 30); July 17 (hatched August 2). A total of 5 cocoons. (3) ? No. 79: May 24 (hatched June 19) ; June 7 (hatched July 5) ; June 19 (hatched July 12) ; June 29 (hatched July 19) ; July 7 (did not hatch); July 17 (hatched August 9); August 1 (hatched); August 21 (hatched). A total of 8 cocoons. (4) 9 No. 117: June 10 (hatched July 6); June 26 (none of the ova developed); July 3 (hatched July 23); July 12 (hatched July 29); July 27 (hatched August 20) ; August 8 (hatched) ; August 28 (hatched September 29). A total of 7 cocoons. (5) 9 No. 118: May 31 (hatched June 5). (6) 9 No. 219: July 4 (hatched July 24); July 28 (hatched August 21); August 21 (hatched). A total of 3 cocoons. (7) ? No. 220: June 29 (hatched July 27) ; July 16 (hatched August 2); July 30 (hatched August 22); August 21 (hatched). A total of 4 cocoons. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 97 The differences in the time of hatching are probably due as much to differences in the thickness of the cocoons as to any such factor as temperature. Care of the Young. — I have seen no evidences that the mother guards in any way either her cocoons or her young. The young make their own way out of the cocoon, making an opening or openings at the margin where the cover is joined to the base; that the mother does not tear the cocoon open for them is proved by the cases Avhere the yoimg emerge at a point that is inaccessible to the mother. Feeding. — This is rather a timid species, the great protection afforded by the tubular portion of the web compensating for want of strength, and they rarely attack prey as large as themselves, and usually avoid much smaller spiders. The prey is killed by biting, and after being sucked dry is usually left hanging in the web. The males especially are frequently eager for water. Agalena naevia Walck. PI. IV, fig. 4. Individuals observed, males: No. 72, captured May 11, moulted May 24, June 7, July 3, July 28, died September 30 (starvation) ; No. 303, captured July 9, moulted August 2, killed by ^ No. 243, August 20. Individuals observed, females: No. 73, captured May 11, moulted May 13, 25, June 7, 29, July 13, August 5, still living; No. 206, captured June 16, moulted June 29, July 27, died in December; No. 243, captured June 22, moulted July 3, July 28, died in November. Moult.— The first moult of ? No. 73 commenced at 11.08 P.M., she hanging with her spinnerets and feet attached to the roof of the cage. The old skin spht into a dorsal and a ventral piece, separated by a horizontal break along the sides of the abdomen, along the sides of the cephalothorax above the legs and below the eyes. She slowly fell out of the split exuvia, with the help of slight convulsive movements of the body and legs. Her legs were completely freed at 11.15, and at 11.23 she hung head down attached to the exuvia only by the spinnerets. At 11.39 she climbed into the web. The new skin is at first much lighter than it becomes on exposure to the air. Mating. — The two cases seen were these : (1) 9 No. 73. cJ^No. 72 was placed upon her web on August 21, at 10.20 A.^I. Both were immovable up to 12.00, when I left them. On my return at 12.40 his fore-legs were touching hers. At 12.47 he shook his body. At 1.20 she flexed her legs close to her sides, he being then upon the web half an inch above her. At 1.26 he seized her from above, his head pointing toward her spinnerets, and tapped her abdo- men several times with his first pair of legs, she motionless. At 1.36 he left her; she remained inert until 1.46, then moved somewhat, 7 98 PROCEEDINGS OF THE ACADEMY OF [Jan., when he quickly turned and faced her and she became again motionless with her legs flexed closely. At 2.06 he pulled her toward and beneath liim, his head facing her spinnerets and just above the pedicel of her abdomen; at 2.08 turned her partially on her side, then applied his left palpus from 2.08 until about 5.00, then applied his right palpus on her other side from 5.00 until 9.40; the number of insertions of each palpus was not noted. At 9.40 he stood over her. At 9.46 she moved slightly for the first time, when he embraced her again and tapped her abdomen rapidly with his fore-legs, got her agaui into copulatory attitude and inserted his left palpus from 9.502- to 9.54. He then withdrew this palpus and cleaned it vigorously with his chelicera until 9.58, when he left her. She quickly rose and followed him, but he escaped. (2) ? No. 243. d^No. 303 was dropped upon her web on August 19, at 3.14 P.M. At 3.34 he placed his first pair of legs upon hers, while he was higher in the web than she. At 3.37 she turned away, then faced him again, and at 3.40 he moved still nearer. At 3.46 he sud- denly grasped her with his legs, and, she flexing her legs to her sides and becoming motionless, he carried her an inch away from the place. At 3.49 he left her still lying inert and walked about, evidently seeking a place on the web favorable for copulation, returning to and standing over her half a minute later. At 3.51^ he seized two of her legs with his chelicera, carried her about, then dropped her. At 3.54 she turned around. At 3.59 he seized her again, placing himself over her with his head directed obliquely toward her spinnerets, she lying partially on one side with her legs flexed, and commenced the copulation with his left palpus. The length of time of each insertion of this palpus was as foUows: 3.59-4.19^, 4.20-4.30, 4.31-4.37^, 4.39-4.46^ 4.47- 4.56, 4.57-5.03, 5.03f-5.08i, 5.09-5.14, 5.15-5.22, 5.23-5.31, 5.31f- 5.4H, 5.42^-5.49^, 5.50-5.55, 5.56-6.01, 6.01|-6.09, 6.09J-6.14, 6.141-6.24, 6.24^6.28, 6.29-6.36, 6.36^-6.56^, 6.57-7.10^ 7.11-7.21, 7.22-7.30i, 7.31-7.40i, 7.41^-7.50^, 7.51-7.58^, 7.591-8.06^, 8.07-8.16, 8.17-8.36^, 8.38-8.39, 8.40-8.54. He inserted the palpus each time by extending it downward along the left side of her abdomen to her epigynum, and inserting the spirally wound process of the palpal organ for its whole length; when inserted, a large sac connected with the ventro-median side of the palpal organ swelled up. The palpus was withdrawn after this sac had collapsed, and the palpal organ drawn through the chehcera before the next insertion. At 8.55 the male changed his position upon her, so as to bring his long axis parallel to hers, lifted her 1 cm. backward, then stood quietly over her. At 1903.] NATURAL SCIENCES OF PHILADELPHIA. 99 8.58 he brought his head obhqiiely over her right side, then apphed the right palpus, in the same manner as he had the left, as follows : 8.58^8.59^ 9.00-9.12, 9.13-9.19, 9.19^-9.24, 9.25-9.30, 9.30|-9.34i, 9.35-9.39, 9.40-9.45, 9.46-9.49^, 9.50-9.56, 9.56^-10.02, 10.02*-10.06, 10.06^10.16, 10.16^-10.23, 10.23^-10.29^, 10.30^-10.39^, 10.40-10.45, 10.45^-10.55, 10.56-11.18, 11.18^-11.31, 11.32-11.46^ 11.47-11.59. Then for very weariness I ceased watching and left them in copula, and next morning found the female eating the male. The female was immovable throughout the whole act, her sternum with its ventral surface downward, but her abdomen twisted on its long axis so as to flex the epigynum up to meet the palpus of the male. His right palpus was used upon her right side, his left palpus upon her left side. The copulation lasted in the first of these cases for 7 hovirs and 47 minutes; in the second case it was watched for exactly eight hours, but lasted longer. In both cases the approach and behavior of the males was alike, and both used the left palpus first. There was nothing in the behavior of the males to indicate a courtship ; there was simply a cautious approach of the male, and after he had found no sign of hostility on the part of the female he quickly seized her, and she was absolutely submissive in his grasp. When he first grasps her he taps her abdomen with his first pair of legs, perhaps an act of subjugation. Not one of the females have made cocoons up to the time of this writing (October 5). Dictyna volupis Keys. PL V, fig. C Numbers of individuals of this species were collected on May 26 upon vines of Ampelopsis and English ivy gi-owing upon the walls of my old homestead, near West Chester, Pa. They had constructed their webs upon the upper surface of the leaves of these plants, and upon most of the webs a male and female were found together, and a number of these were found then in copulation. This species is easily kept in captivity, and on account of their small size I kept them in test-tubes with the mouth loosely plugged with cotton, placing a male and female together in each tube, when they made the webs conjointly; males and females caught together were kept together. Individuals observed, males: No. 130, caught on web of $ No. 131, died June 15; No. 132, caught on web of ? No. 132, died June 24; No. 134, caught on web of ^ No. 135, died in June; No. 136, caught on web of $ No. 137, died June 16; No. 138, caught on web of $ No. 139, died June 3; No. 140, caught on web of Q No. 141, died June 18; No. 142, caught on web of ? No. 143, died July 2 ; No. 144, caught on web of $ No. 145, died June 3; No. 146, caught on web of 9 No. 147, died in June; No. 148, caught on web of ? No. 149, died June 7. Individuals observed, female-'^: No. 131, escaped June 14; No. 1.33, escaped 100 procp:eding.s of the academy of [Jan.. July 2; Xo. 135, died early in September; No. 137, died early in September; No. 139, escaped June 3; No. 141, died about September 1; No. 143, died Au- gust 6; No. 145, still living; No. 147, died about September 28; No. 149, died in September. Mating. — The following were the observed cases worthy of note: (1) 9 No. 133. d'No. 132 was introduced at 12.30 P.M., May 27. Most of the time he spent in spinning upon the web. but sometimes approached and touched her, when she either moved away or else shook her legs tremulously. At 3.40 I was obliged to leave, and on my return at 4.33 found them in copula against the bottom of the test-tube. She lay upon her left side wdth her legs flexed close to her body, her abdomen and cephalothorax in one line. He lay upon his right side with his face pressed against the ventral surface of her cephalothorax and his ventral surface turned toward hers, but not parallel with it, the axes of their abdomens diverging at an angle of mor.> than 45°; his first two pairs of legs embraced hers closely, his posterior pairs were directed backward and braced against the web. This male had lost his right palpus; his left he extended backward along the ventral surface of her abdomen so as to hold the palpal organ inserted in her epigynum. This copulation lasted without interruption up to 5.32. Then after she had made two cocoons, on June 4 and 9 respectively, the same male was found in copulation again at 12.59, continuing so up to 1.48, when he left her, and though he returned several times she met him with open jaws, except once when she approached him with closed jaws, shaking her flexed first pair of legs. Then he proceeded to charge his palpus with sperm. (2) ? No. 135. d'No. 134 was found at 7.35 P.M., May 27, in copu- lation; the embrace was as in the preceding case, except that their bodies were directed upward. Only the right palpus was used, and that was kept continuously inserted ; a large evaginated sac connected with this palpal organ dilated and contracted 10-11 times a minute, each dilation of it accompanied by a slight jerk of his abdomen. He suddenly left her at 8.09, worked his palpi through his jaws, shortly after fed upon a gnat, and no further copulation was observed up to 10.20. On June 13, after she had made three cocoons, they were seen again in copulatory attitude, but his palpi were not inserted. (3) 9 No. 141. d'No. 140 was embracing her at 2.00 P.M., May 31, but his palpi, though pressed against her abdomen, were not inserted. The right palpus was inserted continuously from 2.15 to 3.06. He then moved a few millimeters away, and on approaching her again she rushed at him with opened jaws and chased him the length of the 1903.] NATURAL SCIENCES OF PHILADELPHIA. 101 test-tube. No further copulation was observed up to 4.20. On June 5, at 12.55 P.M., she having made no cocoon in the meantime, they were again copulating, he using his left palpus continuously up to 2.10, then his right from 2.10 to 3.13, after which he ran to the oppo- site end of the test-tube and cleaned his palpi with his chelicera. (4) 9 No. 147. May 26, at 2.00 P.M., c?No. 146 approached and seized her, then inserted his left palpus continuously from 2.04 to 3.02, then at 3.02 his right and kept it inserted up to 4.20, when I was obliged to stop observations; they were separated on my return at 5.25. (5) ?No. 149. At 8.14 P.M:, I saw d'No. 148 come face to face with her, then, each of them tapping upon the web with the first two pairs of legs, they moved backward and forward slowly. This lasted only two minutes, when they both became cjuiet half an inch apart; it was repeated again for a short period at 8.54. On June 1 they were seen in copulation at 4.52 P.M., he inserting his left palpus, and again on June 4, at 11.35 A.jM., continuing up to 11.50, he using his left palpus, and afterward he proceeded to the sperm-induction. In this species the male is shaped quite different from the female, and is a little larger (an unusual condition among spiders). There is no true courtship on the part of the male, for his tapping upon the web with his fore-legs is an act w'hich the female also does. More than one copulation may precede the making of a cocoon; and copu- lation may follow cocooning. The longest time a pair was seen in continuous copulation was 2 hours and 18 minutes, the left palpus here being employed for 1 hour and 15 minutes, then the right for 1 hour and 3 minutes. Sperm-induction. — This was seen twice: (1) d^No. 132 terminated his second copulation with 9 No. 133 at 1.35 P.M. After that I watched him continuously with a lens. From 2.00 to 2.19 he alternated spinning upon various portions of the web with cleaning his palpi with his chelicera. At 2.20 he commenced spinning over a small area, the diameter of which was not larger than the length of his al^domen, and ceased at 2.22. Half a minute later he bent his abdomen downward and discharged from his genital aper- ture a minute drop of sperm, which appeared to adhere to the delicate web beneath him, and at the same moment reached his left palpus (his right was missing) downward and backward Vjeneath his cephalo- thorax and took this drop into the palpal organ. He then remained quiet up to 1.42, when I stopped observations, occasionally shaking his palpus in the air, but discharging no more sperm. 102 PROCEEDINGS OF THE ACADEMY OF [Jan., (2) cJ'Xo. 148, on June 4, ceased his copulation with ? No. 149 at 11.50 A.M. About 4 minutes afterward he started spinning on the glass wall, not using his calamistra, and made a close but scarcely visible silk sheeting at one place. This continued up to 12.05 P.M., when he stood over it, shaking his palpi and rubbing them against his chelicera. At 12.10 a minute whitish globule of sperm fell from his genital aperture upon the silk sheeting, rolled forward upon it (i.e., toward his head), and was immediately taken into one of his palpal organs, and that palpus was then shaken repeatedly many times. After that he remained in the same position, flexing his palpi and pressing them against the silk sheeting. But there was no further discharge of semen, though I watched him continuously with a lens. He spun again in another place from 12.20 to 12.27, then went off to feed upon a fly. The peculiarity of the sperm-induction in this species is its rapidity, and the fact that the palpi are not placed beneath the silk layer made to catch the drop of sperm. Cocooning. — This process was not seen, most of the cocoons having been made in the early morning. Only the conclusion of it was noticed in a couple of cases, i.e., the finishing of the cover of the cocoon; silk was applied that had been drawn out by the calamistra from the cribrellum. And female No. 131 was seen standing over a flat circular disk of silk one day at 12.00 P.M., and sometime before the next day she oviposited upon it and spun another disk as a cover ; this observa- tion shows that this species, as other spiders, makes her cocoon of a base and a cover. The number of cocoons made, and their dates, were as follows : (1) 9 No. 131 : June 4, June 7. 2 cocoons. (2) ?No. 133: June 4 (hatched June 23); June 9; June 17; June 20. 4 cocoons. (3) ? No. 135: June 8 (hatched June 22); June 9; June 13; July 5; July 8; July 9; July 11; July 15; July 31. 9 cocoons. (4) ?No. 137: June 10 or 11 (hatched June 24); July 3; August 3; August 19. 4 cocoons. (5) 9 No. 141: July 3; July 7; July 13; July 26. 4 cocoons. (6) 9 No. 143: July 4; July 24. 2 cocoons. (7) 9 No. 145: July 10; July 15; July 24; July 30; August 11. 5 cocoons. (8) 9 No. 147: June 4 (hatched June 18); June 6 or 7; June 17 (hatched July 3); June 22; July 4; July 18; July 27. 7 cocoons. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 103 (9) ?No. 149: June 14 (hatched June 29); July 16; July 19; July 23; July 29; August 1. 6 cocoons. The quick succession of these cocoons will be noticed, sometimes co- coons being made upon consecutive days. This is because only a few eggs (perhaps never more than 12 or 15) are laid in any one cocoon. The base of the cocoon is a flat disk, generally broader than the more or less arched cover; the cocoons are pure white. In the test-tubes most of them were placed against the glass, others built upon the web; sometimes they were arranged in rows, one placed upon another. Care of the Young. — The mother appears not to remain by her cocoons nor to guard them in any way. The young make their own way out of the cocoons, emerging at one point between the base and the cover. Feeding. — Relatively to its size this species is the most forward in attack, attacking large insects, such as blue-bottle flies, which many larger spiders flee from. It evidently trusts to the great strength of the web, in which large insects become helplessly entangled. The little spider rushes immediately at the entangled victim, seizes hold of it with his jaws without spinning upon it, and keeps his hold though violently shaken by the prey. Theridium tepidariorum C. Koch. PL IV, fig. 2. Individuals observed, males: No. 7, captured April 29, killed May 19; No. 8, captured April 29, moulted the same day, escaped the same night; No. 13, captured May 1, escaped May 3; No. 19, captured; May 2, killed May 16; No. 20, captured May 2, killed by $ No. 17 May 18; No. 71, captured May 8, killed May 23; No. 77, captured May 13, moulted May 15, died (of starvation probably) June 3 ; No. 81, captured May 14, killed May 18; No. 115, captured May 24, killed next day; No. 116, captured May 24, killed next day; No. 123, captured May 26, died June 7; No. 125, captured May 26, died June 11; No. 126, captured May 26; No. 127, captured May 26, died June 6; No. 155, captured May 30, died June 29; No. 217, captured June 19, died June 29; No. 254, captured June 26, killed (by $ No. 253) July 10; No. 319, captured July 14, died August 6. Individuals observed, females: No. 6, captured April 29, died June 19; No. 16, captured April 29, died about September 1 ; No. 17, captured April 29, moulted May 9, killed May 30; No. 18, captured May 2, died (apparently starvation) June 2; No. 26, captured May 2, killed May 8; No. 76, captured May 13, died (starvation) June 3; No. 124, captured (on the same web with (^ No. 123) May 26, died July 3; No. 129, captured May 26, died July 29; No. 154, captured May 30, died about September 1; No. 156, captured May 30, died July 23; No. 202, captured June 11, died July 29; No. 253, captured June 26, died about September 1 . Moulting. — This was observed in one cJ' only; the process lasted about 104 PROCEEDINGS OF THE ACADEMY OF [Jan., half an hour, the spider slowly falHng out of its old skin, and the legs were quite colorless immediately after the moult. Mating. — The method of observation was to keep a female in a cage until she had made a web there, then to drop a o^, handling him as gently as possible, upon her web. The introductory steps of the mating are as often made by the female as by the male, and she often shows quite an insatiable eagerness, even sometimes leaving food to approach the male. As soon as the male commences to move upon her web she recognizes him as a male of her own species, and, when she is eager, commences immediately to signal to him, both spiders being on the lower surface of the web and upside down (the usual position). The female hangs to the web with the third and fourth pairs of legs, and shakes the longer second and first pairs vigorously and spasmodically in the air (when those legs are not attached to web lines), otherwise with them she shakes web lines to which they are hooked. This "signalling" is a sign of eagerness on the part of the female, and so far as I have observed she makes it at no other time than when she is eager and notices the approach of a male of her own species. There are individual differences in the mode of signalling, as well as differences in accord with the degree of eager- ness of the female ; sometimes a female signals without moving from her original position, sometimes with the signalling she moves by short steps toward the male. When she is not eager she either remains motionless, or else rushes hostilely toward the male as at an object of prey; in both cases the male makes no advances, and when she is markedly aggressive he escapes by dropping from the web. The whole attitude of the male is that of combined timidity and eagerness ; he is much smaller than the female and upon a foreign web, and usually acts with great caution. Very freciuently he will climb about the web for a greater or longer period, the female all the while signaling, before he approaches her; often he approaches and touches her several times and each time rapidly withdraws again ; more rarely, the male responds quickly to the signalling of the female and copulates within a minute of time. Sometimes the male, while moving about on the web, on coming across a break in it, will pause to mend the break before ap- proaching the female. The male shows his eagerness by a spasmodic jerking of his abdomen. He tests the eagerness of the female, and finds her position on the web, by grasping with the claws of his first pair of feet the web lines that she is shaking by her signalling, and by drawing these web lines taut he feels her movements all the more dis- tinctly; he approaches gradually nearer her, guided by her signalling, and finally makes a short rush toward her. 1903 ] NATURAL SCIENCES OF PHILADELPHIA. 105 In this species the courtship, if the expression of eagerness may be so called, comes evidently much more from the female than from the male; and that the male is not able to satisfy the female is shown by the fact that the latter continues her signalling for some time after the copulation. The mode of sexual recognition is entirely by touch, b}^ the tension of the web lines. The copulation I have observed many times. Both individuals being on the lower surface of the w^eb (in small cages the female spinning a horizontal web about half-way between the floor and the roof), on the close approach of the male the female hangs to the web by the claws of her third and fourth pairs of legs, her other legs hanging unattached, so that her cephalothorax lies beneath her abdomen. The male places himself parallel to her, his ventral surface opposite hers, his cephalo- thorax also above her abdomen, his position in the web higher so that liis head is opposite her epigynum, embracing her closely with his second and third, sometimes also with the fourth pair of legs, while his long first pair of legs beat about in the air. His palpi are applied to her epigynum generally both at the same time ; sometimes only one palpus is employed during one copulation. The length of time of the application of the palpi is rarely 15 seconds, usually less than 10 seconds, and all this time the female remains perfectly motionless. The copula- tion completed, the male hurries away, or else falls down out of the web and remains for some seconds motionless, his energies evidently spent. After copulating he cleans his palpi with his jaws. Sometimes the male has difficulty in inserting his palpi into the epigynal orifices, and tries ineffectively many times while the female remains motionless. The copulation is so soon over that I have not been able to determine the exact action of the palpi during it. This species copulates very frequently, and there are several matings preceding the making of each cocoon. The female appears always eager except in the time of full pregnancy, and may mate with numer- ous males, as the following data show : (1) ? No. 6 copulated with c?No. 7 at 7.53 P.M., April 29, and this c? showing no more desire was removed at 9.25. Put him with her April 30, copulation at 1.43 P.M. Again on April 30, at 11.12 P.M. Again at 7.44 P.M., April 30, half a minute after he was put in. On May 3 I put him with her again for two hours, but no copulation. On May 4, copulation at 9.20 A.M., five minutes after he was put in; and again at 8.39 P.M. and 8.43 P.M. May 5, put him in again at 7.59 P.M., copulation within half a minute, and a second one at 8.01. On May 6, when I put him upon her web, she acted hostilely. May 8. 106 PROCEEDINGS OF THE ACADEMY OF [Jan., copulation at 5.24 P.M., half a minute after his introduction. On May 11 she was hostile to him, as also on May 13; at this date she was very large with eggs, which she laid May 19. May 24 I introduced c?No. 115 at 4.51 P.M., he copulated immediately; he was removed at 4.05, and c?No. 116 placed with her, copulating at 4.07^ and 4.08. May 25 I put ind'No. 116 from 10.10 to 11.00 A.M., but he avoided her. though she made amatory signals; and cJ^No. 115 behaved similarly on the evening of the same day. May 27, d^No. 125 copulated at 2.50 and 2.53 ; and the same 6" at 8.43 and 8.45 P.M., May 28. On May 28 put in d^No. 128, copulation at 8.57^ and 8.58 P.M. May 29, put in d^No. 125, copulations at 6.00, 6.01, 6.02 P.M.; and he copulated with her on June 1, at 5.30^ and 5.31 P.M. June 2, d^No. 126 was placed in her cage, but he showed no eagerness and was removed at 9.25 P.M. ; then d^No. 125 was put in, and copulation followed at 9.32^. June 30, put in d^No. 125, but no copulation ensued, although she was eager. June 17, put ind^No. 215 at 11.00 A.M., and two copulations followed within 3 minutes. (A total of 27 observed copulations.) (2) ? No. 16 copulated with d" No. 13 at 7.55, 7.56, 7.56+, 7.58, 8.00 and 8.04^ P.M., on April 29, the d^ after each copulation moving half an inch away from her and drawing his palpi through his jaws. May 4 I introduced d^No. 19 from 7.35 to 9.40 P.M., but there was no copula- tion. On May 5, d'No. 19 copulated at 8.14, 8.18 and 8.19 P.M. ; and the next day at 7.39 P.M., a minute after his introduction. On May 8 d^No. 19 copulated with her at 5.33 and 5.37 P.M.; and on May 13 at 4.22 and 4.23 P.M. On May 14 I introduced d'No. 7, on May 15 d'No. 19, on May 16 d^No. 19, but she was hostile on each occasion (she oviposited May 25). On May 25, at 10.10 A.M., just after the comple- tion of the cocoon, d^No. 115 was introduced; she first rushed at him, then held herself in copulating attitude, then became again hostile. May 27 I introduced d'No. 127 at 3.10 P.M., she rushed at him and started to enshroud him, but when he touched her with his legs she ceased, and copulation followed at 3.11 and 3.12. May 28, d''No. 127 was put in at 9.08 P.M., copulations followed at 9.09, 9.09^, 9.10. May 29, d'No. 127 copulated at 5.27 P.M. ; but when d" No. 128 was put in his place at 5.30, she chased him hostilely. June 1 I put in d^No. 127, but she acted hostilely (perhaps from hunger this time, as she had not been fed for three days). June 2 I introduced d* No. 127 while she was feeding, and she chased him from the web. June 7, d^No. 182 was introduced, but though she signalled eagerly he did not mate. June 15 (a cocoon was built by her the preceding day) I introduced d^ No. 182, copulation ensued at 7.25 and 7.27 P.M. (A total of 22 observed copulations.) 1903.] NATURAL SCIENCES OF PHILADELPHIA. 107 (3) ? No. IS was put into the cage of cJ'No. 8 on M&y 2, but she chased him; then cJ' No. 19 was introduced and copulated with her at 11.27, 11.29, 11. 29^ and 11.31 A.M. Two Theridiids of a smaller species were put in at 12.00 M., so that five spiders were now together. 9 No. 18 sometimes would make amatory advances to c^No. 8, but he (evidently not mature) fled from her. Several times she came into contact with (?No. 19, but after a few excited advances he left her each time. She always acted hostilely to the individuals of the smaller species. She copulated with 6^ No. 19 at 12.42; this 6^ later made a rush at the smaller d^ No. 8. The latter subsequently escaped, and I removed c^No. 19 on May 4. On May 8 a freshly caught large d^was introduced, but she seemed frightened by him. On May 9 she copulated with c?No. 71 at 7.55 P.M.; he was left in her cage until May 23, and copulations were noticed at 9.15 P.M., May 13, and 9.15 P.M., May 15. (A total of 8 copulations observed.) (4) ? No. 76. cJ^ No. 77 was introduced on May 13, but she remained immovable, and was hostile to him next day, as she was to cJ'No. 85 on May 15, and to d'No. 77 on May 16. May 18, d'No. 77 copulated at 6.59 and 6.59^ P.M. (Only 2 copulations.) (5) ? No. 124 copulated with d' No. 123 on May 27, at 2.17, 2.22, 2.27 P.M., besides some unsuccessful attempts. The same d^made several unsuccessful attempts on May 28, and one successful one. May 29 I introduced him while she was feeding, she left her prey and signalled vigorously, copulation at 5.49 P.M. May 30, copulations with the same cJ^at 4.44 and 4.49 P.M.; and at 4.55 P.M. d^No. 155 was intro- duced, but he would not answer her amatory signals. On June 1 d^No. 123 made protracted but unsuccessful attempts to copulate at 5.43, 5.46, 5.47 and 5.47^ P.M. June 2 she copulated with d'No. 128 at 10.28^, 10.29 and 10.32 P.M. June 8 I put in d'No. 128 again at 1.36 P.M., while she was eating a beetle; she signalled immediatel}' while remaining close to the beetle, he went toward her and mistakenly tried to copulate with the enshrouded beetle, but ultimately copulated with her at 1.40 and 1.42, June 12 I put in c^No. 128 again, copula- tions followed at 8.59 and 9.16 P.M.; I left them together, and next morning found she had eaten him . (A total of 23 copulations observed , of which a number were unsuccessful.) (6) 9 No. 129 copulated with d" No. 123 at 8.31 P.M., May 28; later he attempted to insert his palpi but in vain, she then became aggressive. June 2 she copulated with c?No. 155 at 8.37 and 8.43 P.M. June 7, with d^No. 155, at 8.24 P.M., half a minute after his introduction. June 15, with the same d*, at 7.23 V.'Sl. (A total of 5 observed copu- lations.) 108 PROCEEDINGS OF THE ACADEMY OF [Jan., (7) $ No. 156 copulated withe? No. 123 at 12.10. 12.12, 12. 12^, 12.13 P.M., June 3. June 19, three times with d^No. 217. July 15, at 3.30 P.M., with (5^ No. 319. (A total of 9 observed copulations.) These observations show that while the copulations are of brief duration, generally less than 10 seconds, they are frequently repeated. The number must be larger in natural conditions than my figures show for caged individuals, since in order to count the number I was obliged to remove the males during those periods when I could not watch them. Even when there are one or more cocoons in her web the female mates as freely as before oviposition, and their presence does not in any way seem to retard her mating impulses. Sperm-induction. — The following was the only case observed. d'No. 319, after copulating with $ No. 156 at 3.30 P.M. on July 15, was watched carefully with the hope of seeing this process. 'He left the female and went to a corner where he remained quiet, hanging with the ventral surface^ uppermost and the cephalothorax a little lower than the abdomen. At 4.35 P.M. he flexed his abdomen slightly on its pedicel so as to elevate its apex, deposited from his genital aper- ture a small drop of sperm upon a line of the web, then applied the jialpal organs alternately to this drop. This process continued until 4.39 P.M., when he was driven away from the place by the approach of the female. The droplet of sperm had been only slightly diminished in amount, so that probably the induction into the palpal organs takes a much larger time. I watched him for half an hour more; for a while he shook his palpi slightly in the air, but ceased this action and did not emit any more sperm. ! Cocooning. — The brown ovoid or subglobular cocoons are very familiar objects in barns and cellars where this species is most fre- C}uently found. In captivity they make their cocoons usually in the early morning, completing them before 8.30 AM., in one case as late as 10.00 A.M. The process is as follows: 9 No. 16 was observed commencing her first cocoon at 8.15 A.M., May 25. She was making a fluffy ball of loosely curled white silk, barely a third the diameter of the finished cocoon, suspended by a thread from the roof of the cage. She worked by hanging in the web with her first pair of legs, using the legs of the fourth pair in alternate action to pull out from her spinnerets and apply to the silken mass the white thread, and occasionally helping with the third pair of legs. At 8.23 J she placed her epigynum close to the lower surface of the silken ball, and there issued from her genital aper- ture a large, yellowish, viscid drop of fluid in which ova could be 1903.] NATURAL SCIENCES OF PHILADELPHIA. 109 indistinctly seen ; this drop was not very transparent, so that I could not determine whether the drop was first exuded and then the ova rolled into it, or whether they were discharged together, but appear- ances were more in favor of the last supposition. This drop was very viscid and of thick consistency, holding a spherical shape and enclosing the ova, and adhering firmly to the silken ball. At 8.27 she completely freed herself from the drop, and then commenced to spin over it. This was accomplished by hanging with the first and second pair of legs from the web, the legs of the fourth pair in alternate use applying the thread to the surface of the drop ; while the third pair of legs and the palpi were employed in concerted action so as to revolve the whole mass on its suspending silken thread. The spinnerets at this stage were not applied directly to the cocoon, From 8.27 until 8.55 she was on the left side of the cocoon, revolving it slowly from right to left; from 8.55 to 8.58 on the right side, then turning it in the opposite direction; from 8.58 until 9.00 on the left side of it, revolving it again from right to left. At this time the cocoon was of nearly its final size, almost spherical, its diameter a little greater than the length of her body. The egg mass and the original silken ball were completely covered and hidden by closely matted brownish silk. From 9.00 until 9.17 she crawled slowly over its surface, pressing against it with the tips of her palpi used alternately, evidently to mat down and smooth the whole, the feet being used simply to hold on with. At 9.17 she started applying new anchoring threads to the cocoon, this process continuing until 9.25; this was done by applying the spinnerets directly to the cocoon, then carrying each thread so fastened to an adjacent part of the web; about 15-20 such supporting threads, attached to different parts of the cocoon, were made. She rested until 10.05, then made a few more supporting lines. 9 No. 202 likewise commenced cocooning by making a ball of curled threads, as did No. 16, her cephalothorax placed above her abdomen during the process. This continued up to 8.28 A.M., when she held her epigynal aperture against the lower surface of the ball and ovi- posited upon it at that point; the oviposition lasted from 8.28 to 8.31, the ova coming slowly out of the genital aperture one at a time and in- cluded in a viscid drop. At 8.32^ she commenced spinning again, covering first the lower surface of the mass of ova, then the sides, the cocoon becoming regular in form by her revolving it. This continued until 9.25 without interruption, when she stopped spinning, and instead slowly revolved the cocoon (now brown in color) and kneaded its surface with her palpi; this continued until 9.45, when she bit loose 110 PROCEEDINGS OF THE ACADEMY OF [Jan., certain web lines around the cocoon, and fastened new supporting lines to it. 9 No. 154, after making 3 normal cocoons, all suspended in her web, made a mistake in the construction of her fourth one. On July 24, at 8.13 A.M., I found her finishing the base (flviffy ball) of the fourth cocoon; but owing to some accident this was not a rounded mass of silk, but a hollow cone hung from its apex by a thick thread ; the form had probably become changed by her weight upon it. From 8.20 until 8.23 she clung to this base, pressing her epigynum several times against its lower opening, and from 8.23 to 8.24 she oviposited upon it. But her weight upon the silken base had gradually unraveled the latter, so that it becoming pulled out, the egg mass sank with it to the floor, where all adhered to some dead flies. From 8.24 until 8.40 she made several ineffectual attempts to raise it into the web, then she left it and went to cocoon No. 3, spun a cap of silk first around its lower surface, then upon its sides, then kneaded this new covering with her palpal tips. Evidently she had the impression that she was finishing the cocoon already begun, and seemed to have forgotten the neglected egg mass. At 10.05 she returned to the latter and fed on the eggs, probably confusing them with the dead flies to which they adhered ; at 10.10 she pulled excitedly at the neglected base, as if trying to raise it, for 6 minutes, but then left it on the floor and did not return to it again. Thus she had not been able to rectify the misshapen base of the cocoon, and under the impulse to make a covering had made this covering around an already finished cocoon. In all these cocoonings, then, a base of a spherical mass of loosel}' curled silk is first formed, the ova (included within a viscid drop of fluid) deposited upon its lower side, then the whole covered with silk, finally this covering kneaded mth the ends of the palpi. The reason why the external surface of the cocoon is brown in color I cannot say ; possibly the silk secreted last maybe different chemically — i.e., proceed from different spinning tubes. In one case a complete cocoon was made without oviposition. A succession of cocoons are formed by the same individual, the num- ber of those made by my captives with the dates of th(:ir construction being as follows : (1) $ No. 6: May 19 (the ova did not develop). (2) 9 No. 16 : May 25 (hatched June 28) ; June 14 (hatched) ; July 7 (eggs did not develop); July 14 (eggs did not develop); July 24 (hatched); July 31 (only a few of the eggs hatched); August 21 (hatched). 7 cocoons in all. 1903.] NATURAL SCIENCES OF PHILADELPHIA. HI (3) Q No. 129: June 2 (hatched June 14); June 24. 2 cocoons. (4) 9 No. 154 : June 3 (hatched June 15) ; June 8; June 24 (hatched) ; July 24 (this was the incompleted cocoon described above, the eggs shriveled up); August 6 (eggs did not develop); August 15 (only a few of the ova hatched) ; September 1 (hatched). A total of 7 cocoons. (5) ? No. 156: June 28 (this cocoon contained no ova). (6) 9 No. 202: June 14 (hatched June 28); June 24 (hatched July 8) ; July 3 (hatched July 14) ; July 12. A total of 4 cocoons. (7) 9 No. 204: July 14, the ova after a few days were shaken out of this cocoon into a test-tube to see if they would develop, and they all hatched on June 26. (8) 9 No. 253: July 24 (hatched August 4); July .30; August 6; August 16; August 27. A total of 5 cocoons. Probably the rapidity in the succession of the cocoons is directly dependent upon the richness of the food supply. The time of hatching would be dependent, among other factors such as temperature, upon the thickness of the cocoon covering through which the young have to escape. Care of the Young. — There seems to be solicitude for neither the cocoons nor the young, though the mother will not eat the latter. The mother does not guard the cocoons, and she stays near them simply because they are generally placed in that highest portion of the web where she is accustomed to dwell. Sometimes on the introduction of a fly, the mother will enshroud it alongside of a cocoon, and after sucking it dry, will cut it, with the cocoon, loose from the web; such fallen cocoons are not lifted again into the web. The young, on hatch- ing, move in the direction of the strongest light, whereas mature indi- viduals avoid the light ; this instinct would serve to scatter them from the maternal home. Feeding. — Living prey is recognized entirely by its pulling upon the web lines; and this species is one of the boldest of our local spiders, attacking without jiesitation insects and spiders much larger than itself. That is, the females do, but the males are much more timid. A rush is made toward the struggling insect, a line of silk attached to it, and then the spider, hanging by her first pair of legs, applies the issuing silk thread to the victim's surface by the very rapid application of the three last pairs of legs. Not until the victim is closely enshrouded does the spider bite it, and she enshrouds just up to that point when the victim is rendered motionless by the envelope of silk. Often the enshrouded prey is carried to another part of the web to be eaten ; and after being sucked dry is bitten loose from the web and allowed to fall out of it. il2 PROCEEDINGS OF THE ACADEMY OF [Jan., Teutana triangulosa Walck. This is locally quite a common species in houses, but much less abun- dant than Theridium tepidariorum. Tlie following individuals were kept in captivity: cJ* No. 173, captured June 4, killed by Q No. 70, June 19; 9 No. 70, captured May 8, killed by accident August 15; ? No. 153, captured May 28, died July 22; ? No. 176, captured June 5, still living. Mating. — Four different males of Theridium tepidariorum were placed upon the web of 9 No. 70, but she was hostile to each. On three different occasions cj' No. 173, of her own species, was placed upon her web; she signalled to him by pulling upon the web lines, shaking her first pair of legs, but he avoided and was finally eaten by her. Her mode of signalling was like that of Theridium tepidariorum. Cocooning. — This process I saw only once. 9 No. 153 was seen at 9.05 A.M., June 15, to have spun a small ball composed of looped threads suspended in the web; she continued to add to it until 9.10, holding to the lower surface of the ball ("base" of the cocoon) with her ventral surface turned toward it, raising her spinnerets to apply them to the ball, dropping them, then raising them again for another appli- cation, the length of the "stroke" accounting for the looseness of the threads. At 9.10 she placed her genital aperture close to the lower surface of the silken ball, and deposited there a globular, viscid mass enveloping ova; this oviposition lasted just half a minvite. Then she immediately began spinning over the egg mass, laying down long threads upon it, making the covering spherical by revolving her body around the cocoon, whereby she first supported herself by the sur- rounding web lines, later by hanging to the cocoon itself. She stopped spinning at 9.21. The cocoons of this species are snowy white, spheri- cal, and peculiar in that they are so loosely constructed that the yel- lowish egg mass may be seen through them. In natural position the cocoons are hung high in the web, generally attached to some object (such as a board) that rpofs over the web; in captivity some of the cocoons were attached to the roof of the cage in conformity with this habit, but as often they are placed lower in the web. In natural conditions, also, the cocoons are generally placed close together, in that part of the web where the spider spends most of her time. The number of cocoons made by my captives, and the dates, were as follows : (1) 9 No. 70: May 21, June 3, 15, July IS, 26, August 2, 13. A total of 7 cocoons. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 113 (2) $No. 153: June 2 (hatched June 27), June 8 (hatched August 14), June 15 (hatched July 17). A total of 3 cocoons. (3) 9 No. 176: June 15, 26 (hatched July 16), July 10 (hatched July 30), 16, 27, 31, August 5, 15, 21, 28. A total of 10 cocoons. Care of Young, Feeding Habits. — I found no indications of maternal guarding of the cocoons or young, but the mother appears indifferent to both. It is the habit of this species, after sucking a victim dry, to cut it loose from the web, and at such times a cocoon is sometimes cut loose also, and then not raised again into the web. This species is very courageous and rushes at large flies, quickly enshrouding them with silk, which is applied with the fourth pair of legs ; sometimes the struggling victim chances to be in contact with a cocoon, and then the spider frequently enshrouds both together. The young cut their own way out of the cocoons. Steatoda marmorata (Hentz). One male and five females were kept in cages. Mating. — d^ No. 46 was placed upon the web of a female, both re- mained motionless for a while, then she killed him. Another female killed an introduced male of S. borealis (Hentz), Cocooning. — This was seen only once, and only the end of the pro- cess observed. At 7.30 A.M., May 28, I found that a female had ovi- posited upon a mass of silk, and was then engaged in spinning over it. The spider held to the cocoon with her feet, and spun upon it (not using the legs to apply the thread) by attaching the spinnerets at one point, elevating them, and applying them again close to the first point of attachment, thus forming looped threads; in one minute I counted 22 such applications of the spinnerets. The spider in spinning slowly rotates around the cocoon. The cocoon is suspended in the web, white in color, spherical and rather loosely made. In no case were more than 2 cocoons made. Feeding Habits. — This species is relatively very powerful and coura- geous, and quickly overcomes its victims by discharging upon them a thread that is remarkably viscid and tenacious; minute viscid drops can be seen upon this thread with the naked eye. This is a process of enshrouding that consists, not in WTapping the victim in threads, but rather in pouring out the very sticky secretion upon it. The spider stands upon her first two pairs of legs with her posterior end toward the object of prey, and applies the viscid thread to the latter with her other legs. 8 114 PROCEEDINGS OF THE ACADEMY OF [Jan., Linyphia clathrata Westr. Moult. — At 6.40 P.M., May 7, I dropped a male of this species on to the floor of a cage, and after a few minutes he fell upon his left side and commenced to moult. His palpi and all four pairs of legs were held straight at right angles to his ventral surface. The whole process lasted somewhat less than five minutes. The old skin split by a hori- zontal break, just below the eyes anteriorly, and on the sides just above the legs and along the sides of the abdomen; thus it opened in two pieces, a dorsal and ventral, which remained connected together by a narrow §trip close to the spinnerets. His repeated jerking flexions of the cephalothorax gradually freed first his cephalothorax, then his legs and abdomen. After the moult he rose to his legs, but was unable to walk until the expiration of 16 minutes. Pholcus phalangioides Fues-sl. PL IV, fig. 3. Individuals observed, males: No. 87, captured May 16 on the web of $ No. 86, killed by 9 No. 82, May 27; No. 88, captured May 17, killed by female No. 12 the same day; No. 175, captured June 5, died June 7; No. 180, captured June 7, escaped June 16; No. 245, captured June 23, died June 26. Individuals observed, females: No. 9, captured April 29, still living; No. 12, captured April 29, died August 15; No. 15, captured May 1, moulted May 24, still living; No. 82, captured May 15, moulted May 22, still living; No. 86, cap- tured May 16 (on the same web with cJ* No. 87), killed May 31, and immature at death; No. 174 captured June 5 (with a cocoon), died July 5; No. 179, cap- tured June 7. died June 25 (from starvation) ; No. 181, captured June 7, died August 7. Mating. — The following data include the more important observa- tions : (1) ? No. 9. On June 5, at 12.49 P.M., after she had been well fed, 6^ No. 175 was placed on her web. He first remained quiet, then touched her with his legs; at 12.56 she reached toward him and he fell from the web. At 1.06 he climbed into the web, and she moved toward him. At 1.25 he hung below her and touched her fore-legs with his, when she drew her legs away, he continuing to stretch out his first pair of legs toward her. She had in the meanwhile drawn up to herself a former enshrouded victim, and this intervened between them when, at 1.43, he tried to embrace her, so that he removed it and touched her head with his chehcera. She hung perfectly motion- less in her usual position, her abdomen above her cephalothorax and directed vertically upward. The position of his body was the same, but his legs were outside of hers. At 1.47 he inserted both palpi at once, and hung to her by them, his weight puUing her abdomen into the horizontal plane. The position of the two in copula was then as 1903.] NATURAL SCIENCES OF PHILADELPHIA. 115 follows : The female with her abdomen horizontal and ventral surface down, her cephalothorax flexed at an angle downward; thee? below her, his cephalothorax about vertical and his head almost touching the ventral surface of her abdomen at the anterior end of the latter, his abdomen flexed at more than a right angle with his cephalothorax ; the long axis of his abdomen thus made a wide angle with the long axis of hers, like a V. Thus they continued motionless until 2,19, when he dropped to the floor, and when he attempted to climb into the web again she chased him so vigorously that I removed him. On June 7, at 7.55 P.M., c? No. 180 was put in; they embraced in copu- latory attitude for a minute, but she moved her legs, and he left her and went to another part of the web. At 8.17 she tried to enshroud him, so I separated them. On June 14 I put in cJ^ No. 180 again, at 4.36 P.M. ; he remained quiet for six minutes, then climbed up and touched her, whereupon she made her shaking motion. He then hung beneath her, shaking the web very gently by a swaying of the whole body, she responding in a similar way. At 4.47 she rushed at him, and I removed him. (2) 9 No. 12 killed and ate mature 6" No. 88 on May 17, and an immature c? on May 24, and made her first cocoon on May 27. On June 23, c^No. 245 was introduced at 12.12 P.M. (the cocoon had hatched on June 15). At 12.21 he started to move with long pauses, while she was above him in the web and testing his movements by draAving the web Hues taut. At 12.28 he inserted his palpi in her epigynum for a second, then fell to a lower position upon the web. At 12.30 he made several attempts to insert both palpi, and succeeded half a minute later. The position of the two was as in the preceding case, she hanging in the web by her posterior three pairs of legs, he hanging to her by his inserted palpi, his anterior three pairs of legs loosely embracing hers (and their claws hooked to web lines), his fourth pair of legs directed behind him and also hooked to web lines. The palpi were inserted by pressing them in this flexed position along the ventral surface of her cephalothorax backward to her epigynum. When inserted, a large vesicle or sac was evaginated from the lateral side of each palpal organ. I watched the copulation from 12.30 to 12.55, when I was obliged to leave, and on my return at 1.25 found them separated, he feeding on a beetle which she had previously enshrouded. (3) ? No. 82. I put d" No. 87 with her from May 17 to Jlay 22, when he w^as repeatedly observed to make advances toward and to touch her, but she acted slightly hostile to him and no copulation was 116 PROCEEDINGS OF THE ACADEMY OF [Jan., seen. She moulted on May 22. On that day (he had remained all this time in her cage) I watched a curious kind of courtship between them from 8.00 to 9.40 P.M. They hung opposite each other in the web, he by his first pair of legs, she by her first and second, reaching out and parrying with their other legs. His head pointed directly down- ward, her body in a horizontal position with the dorsal side down. During this process, wliich continued with few interruptions during the period of observation, he changed his position only slightly, while she moved about somewhat. Sometimes he made a short advance toward her, then not holding his palpi directed anteriorly (as usual) but with their stalks bent ventrad, so that the large palpal organs were bent upon his cephalothorax ; at each such attempt at copulation the striking of her legs against him seemed to frighten him, and he with- drew. Both became quiet at 10.00. If this was a courtship, she appeared more active in it than he. On May 23, 24 and 25 I watched the pair at intervals, but saw no copulation. But at 10.20 P.M. on May 26 I found them in copula; they certainly had not been more than 15 minutes in that act. This. copulation lasted without interruption to 11.31. Then he climbed to the bottom of the cage, when she actively chased him, and finally contented herself with feeding upon a former enshrouded victim. The attitudes of the two were as in the preceding cases. Both palpi were continuously inserted at once, and neither withdrawn, though each at intervals underwent a slight lateral move- ment; they were so firmly fixed in her epigynal apertures as to bear the greater part of his weight; at each side of each palpal organ was clearly visible an evaginated, rounded sac, which collapsed after the withdrawal of the palpi. This male had lost his third pair of legs; with his fourth pair directed caudad he hung to the web ; the left foot of his second pair was hooked to the tarsus of the left leg of her second pair; the right foot of his second pair was hooked to the tarsus of the right foot of her fourth pair; the left tarsus of his first pair rested against the left tarsus of her fourth pair; and the right tarsus of his first pair rested upon the tarsi of her first and second pairs. Next day she killed this male. (4) While her young were scattered upon her web, I put in cJ' No. 245 to 9 No. 174. At the end of five minutes she rushed at and tried to enshroud him, when I separated them. From these observations it is apparent that there is no real court- ship on the part of the male; he approaches her with great caution, tentatively touching her with his legs, and only then attempting copu- lation when she remains immovable under his touches. In all the cases 1903.] NATURAL SCIENCES OF PHILADELPHIA. 117 the female attempted to kill the male after copulation, and very fre- quently before; then he escapes by falling from the web, and always approaches her from beneath. The time of copulation varied from 11 to 32 minutes. Cocooning. — The making of the cocoon I have not seen, though I have watched pregnant females for long intervals in the hope of seeing it. Pokrowsky (1899) is the only wTiter who has described this pro- cess. The cocoon is a very thin, white, silken covering to the globular mass of eggs; so thin and flimsy that we can be sure that there is a real silken cocoon only by the white case that is left after the young have escaped from it. The number of cocoons furnished by my cap- tives was as follows, each date given representing the time of making of a particular cocoon: 9 No. 9, June 24 (hatched July 12), July 26, August 30 (hatched September 23) ; ? No. 12, May 27 (hatched June 15), July 19 (this cocoon was dropped to the floor and did not hatch) ; ? No. 82, June 14 (hatched July 4), July 19 (this cocoon was dropped and did not hatch), August 16 (hatched about September 1), latter half of November, (hatched in December); 9 No. 179, June 16 (this was dropped and did not hatch); 9 No. 181, June 23 (this was dropped and did not hatch), July 19 (hatched August 4). Thus none produced more than 3 cocoons, and when a cocoon was dropped it did not hatch. Care for the Young. — As nmiierous writers have noticed, the mother carries her cocoon suspended in her jaws. But it has always been stated that the cocoon is carried until the young hatch. While I have found this to be the case in one or two instances, I have seen more cases where the mother will occasionallv suspend her cocoon temporarily by a line to the web, in order to clean herself and to feed. This is the more remarkable since rough handling does not cause the mother to loose her firm hold upon the cocoon, and her jaws are so firmly imbedded in its lining that, even when she wishes to, she has difficulty in breaking her hold upon it. Perhaps some of these cases where the cocoon falls to the ground are mishaps occasioned by the mother losing her hold upon her cocoon when trying to suspend it upon a thread. That such a suspension is frequent is shown by the following facts: ? No. 9 held her cocoon continuously in her jaws from June 24 up to July 2; on the latter date I gave her a fly, she chased and quickly enshrouded it, then left it and went to the roof of the cage, where she spun a few fine lines, then with the help of her second and third pairs of legs removed the cocoon from her jaws, held it with these legs while she touched her spinnerets to its surface three timeS; then left it hanging on the fine lines thus made, and which were at- 118 PROCEEDINGS OF THE ACADEMY OF [Jan., tached to the roof by her third pair of legs, and went back and fed upon the fly; after this meal she returned and took the cocoon into her jaws again. She repeated this process next day with another fly, and the day following without a fly being put in. ? No. 12 was ob- served to do the same operation once with her first cocoon to secure food, and once with her second in order to clean herself. 9 No. 82 was observed to do so also. 9 No. 181 was observed to do so likewise in order to clean herself, and she, when regaining the suspended cocoon, first seized its envelope with her jaws, then with her legs snapped the supporting lines. The hatching of the young has been observed before, but never very fully described, so that a few notes may be of interest. 9 No. 82 was observed on the afternoon of July 4 holding her cocoon in her mouth, the young slowly unfolding themselves, the most superficial first; at 2.40 all the young had unfolded and hung as a great mass to her jaws; at 6.15 all had dispersed and the empty cocoon was hanging in the web. With 9 No. 174 the process was noted in more detail. At 8.30 P.M. the young were beginning to unfold their legs at the lower side of the cocoon, which the mother still held in her jaws; the more superficial of the spiderlings stretched themselves, and by the enlacement of their legs together, and enlacement with the legs of those that succeeded them, there was attached to the mother's jaws a great mass of young spiderlings, a mass of a diameter fully equalling the length of the mother's body. Next morning early I found all the young, nearly 60 in number, distributed over the web, the mother motionless near them, and lower in the web the empty cocoon. 9 No. 12 acted differently. On June 15, at 2.30 P.M., I found the young spiders beginning to escape from the cocoon, which in this case was suspended from the web, and not from the mother's jaws. At 2.35 she tore away with her jaws a por- tion of the cocoon, which allowed the young to emerge more easily; this she repeated at 2.40 and 2.43, each time avoiding biting the young spiders. By this time about 17 of the young were upon the web, and the mother moved about the young, spinning, and careful not to injure them. At 3.40 I put a large beetle {C haul log nathus sp.) upon her web, which became entangled there and shook the web. For a minute the mother remained quiet, then went to the mass of young which had not yet become scattered, and enshrouded it with silk as she does a victim, but more slowly; then carried it with her jaws. I watched her until 6 o'clock; at intervals she left the enshrouded mass of young hanging on the web, but always returned to it again to take it in her 1903.] NATURAL SCIEXXES OF PHILADELPHIA, 119 jaws, and sometimes tore at the silken shroud with her jaws. At 6.15 she started energetically tearing open the shroud, but all the young did not emerge from this silken envelope until a couple of days had passed. It is difficult to explain this act: did she enshroud the young to protect them from the beetle? Or did she enshroud them mis- taking them temporarily for the beetle? Feeding. — It fills one with admiration to watch this feeble spider, with her w^eak jaws and long hair-like legs, overcoming prey much larger and stronger than herself. It is done by enshrouding the victim with great rapidity. In the operation the spider rushes at the victim, then with cephalothorax up and abdomen down applies with her fourth pair of legs, these legs used in very rapid alternate action and sometimes aided by the third pair, a thread issuing from her spinnerets to the victim, and so within a short time renders it immovable. In case web lines interrupt the free action of her legs, she stops her work to bite these lines loose. The act of enshrouding lasts until the victim is nearly hidden by the thread around it, and completely powerless to move, when the spider carries it hooked to her fourth legs up to a higher place in the web and feeds upon it there. The prey is sucked dry, then cut loose from the web. The agility and strength evinced by this spider is marvelous. The Shaking Habit. — This is the well-known habit this species has of shaking the body so rapidly by a rotary movement, while in the web, as to render itself almost invisible. Certain Epeirids have the same habit. With Pholcus it is an expression of fear, and is resorted to after the web has been severely jarred. Once a female was seen to shake this way after a male w^as put upon her web ; but I never saw such a case repeated, and believe that the female was shocked by some action on my part. It is never resorted to in order to secure prey. Acrosoma gracile Walck. PI. IV, fig. 5. This Epeirid, common in many of the local woods, is remarkable for great sexual dimorphism. The abdomen of the female is hard and corneous, angular and prolonged into spines. The male is much smaller than the female and much more rarely found; his small abdo- men is elongated and without spines. On account of these differ- ences it seemed desirable to observe the mating process. $No. 322 was captured July 23, and died August 11; she made in her cage only the foundation lines of a web, but no orb, and so was unable to catch the insects given her. c? No. 321 was caught on July 23, and died July 28; he also constructed no orb. 120 PROCEEDINGS OF THE ACADEMY OF [Jan., Mating. — On July 27, at 9.27 A.M., the male was dropped upon her web. At 9.31 he touched her, and copulated from 9.32^ to 9.42. She hung ventral surface up by her fourth pair of legs to the web lines, her other legs flexed close to the cephalothorax, her thorax horizontal. The male approached her from in front of her head wdth his long axis also in the horizontal plane, walking upon the upturned ventral sur- face of her cephalothorax and abdomen, when he firmly implanted his outstretched left palpus in her epigynal orifice; then keeping the palpal organ there, he swung his body round so as to bring it on the right side of her abdomen, so that now his head was upmost and a little posterior to her epigynum, and the long axis of his body was approxi- mately parallel to an imaginary line joining her spinnerets with the most posterior spinous processes of her abdomen. Here he held on firmly with his feet to the roughened surface of her abdomen, prac- tically out of reach of any of her legs, and was not dislodged even though part of the time she climbed about as if trying to get rid of him. A swollen evaginated sac connected with the inserted palpal organ was clearly seen in dilation, and he held his palpus inserted until it had collapsed. He then disengaged himself from her, but returned and copulated again, from 9.55 to 10.03, this time placing himself on the left side of her abdomen and using his right palpus. Then he left her again and remained quiet most of the time up to 10.30. At 10.32 he approached and touched her, she followed him and once embraced him lightly with her legs, but he avoided her. After that she climbed about, and whenever she touched one of the web lines which he was holding, by a sudden contraction of his legs he pulled the line taut; this was possibly an amatory signal. I watched the pair continuously up to 1 o'clock, but saw no further copulation nor any preparation for sperm-induction. Epeira labyrinthea Hentz. Of this abundant and beautiful Epeirid only two individuals were watched, c? No. 281 was captured July 2, moulted July 5, died about September 1 (from hunger). 9 No. 280 was captured July 2, moulted July 13 and 18, and died (from hunger) about October 1. This species, as well known, fabricates a labyrinth of lines behind a vertical orb; the female made such a complete web, but the male made no orb. Mating. — After the female had constructed her web and had made 1 cocoon, the male was dropped upon her labyrinth on July 26, at 8.07 P.M. She was in her usual position, just beneath her cocoon, 1903.] NATURAL SCIENCES OF PHILADELPHIA. 121 placed high in the labyrinth. He remained immovable until 8.39, when he began to climb about, alternately advancing toward and retreating from her, spinning threads all the while, she remaining suspended head down near her cocoon, occasionally jerking the wel) lines with her anterior legs (evidently signalling). The male finally made two or three quick motions toward her, withdrew each time, then copulated with her at 8.44 and again at 8.44^. Each copulation lasted barely 5 seconds; she hung head down with legs loosely flexed, like the female of Theridium tepidariorum, her ventral surface turned toward him ; he embraced her legs with his, head downward also, but the copulation was so rapid that I could not determine how he applied his palpi, except that both seemed to be applied simultaneously to her epigynimi. He then left her and went to a corner of the cage, where he vigorously cleaned his palpi and legs up to 8.58; while she remained quiet in the same place. There was no sperm-induction up to 10.00 P.M., when I left them together; he remained upon the laby- rinth of her web up to the time of his death, and the pair seemed per- fectly amicable. There is no courtship, the male finds the female by pulling upon the lines of the web, and her answering pulls are evidently a signal to him of desire upon her part. In natural conditions several males are often found at once upon the labyrinth of the female. Cocooning. — This female constructed five cocoons, on July 26, July 31, August 9, August 19, and August 27 respectively. In accord with the well-known habit of the species the cocoons were placed in a chain one below the other, the latest formed cocoon being the lowest; the cocoons are brown in color, biconvex, rather conical above and more rounded below and irregularly circular in greatest outline. Though the process of cocooning was not seen, it is very probable that the cocoon is formed of a base and a cover, and that the base is placed highest, and for these reasons : First, there is a well- marked line around the equator, showing a union of two parts. Sec- ond, the female, before cocooning, constructs a little tent, in the form of a shallow inverted cone, placed high in the labyrinth, and rests beneath it; this cone is an admirable protection against rain and simlight, and serves to partially conceal the spider. The males appear not to make such nests. Now when a cocoon is made the female uses this conical nest as the base, thickening it by silk, the eggs are laid in its hollow, then a more nearly flat cover of less diameter is spun to close the mouth of the cone. Then she constructs beneath this cocoon a second nest-cone in which she rests, later incorporates it into a second 122 PROCEEDINGS OF THE ACADEMY OF [Jan., cocoon; and so all the cocoons are made (frequently up to the number of 6). Here the nest accordingly is later made a part of a cocoon; probably other Epeirids that make similar silken nests will be found to cocoon in much the same fashion, though I know of one exception (Epeira benjamina Walck.). Sitting at twilight during a heavy down- pour of rain beneath some trees for protection, I saw a female of this species constructing the base of her cocoon on the under surface of a leaf, at a considerable distance from her nest (the nest is a leaf curled up and lined with silk). Feeding. — The female rests below her conical nest supported high in the labyrinth behind the orb, and holds taut with her feet a line joined to the orb. When a fly becomes stuck to the viscid spirals of the orb, she quickly runs along the drop-line to the orb, kills it by biting (without enshrouding it), carries it back to her nest to suck it dry, then usually leaves the carcass hanging in the labyrinth. The males that are upon the web of a female during the mating season also feed upon insects caught in her orb, carrying the victims to outlying parts of the labyrinth. Xysticus stomachosus Keys. PI. V, fig. 7. Individuals observed, males: No. 5, captured April 27, moulted May 15, died June 7 on account of injuries incurred in an attempt to escape; No. 108, cap- tured May 19, escaped May 27. Individuals observed, females: No. 4, captured April 27, moulted May 10, died August 21; No. 109, captured May 19, moulted May 24, still living. Moult. — 9 No. 4 was found in nearly completed moult at 7.40 P.M., the exuvia adhering then only to the spinnerets. She was lying upon the floor on her left side, with her legs stretched out straight, and later fell over upon her back. She did not rise until 8.03, and 5 minutes later chmbed up the wall. This moult has not resulted in any regen- eration of a leg broken off six days before. The old skin was in a dor- sal and a ventral piece separated by a symmetrical horizontal break, as in other spiders. 6" No. 5 was found at 8.20 P.M., hanging head downward from the exuvia, which was fastened to the wall; he did not move until 8.38, and did not break away entirely from the exuvia until 8.43. Mating. — This was studied by keeping a male and a female in the two compartments of a partition cage, removing the partition to see the mating; the spiders showed that they saw each other through the glass partition. c? No. 5 was placed with ? No. 4 on a number of occasions prior to May 10, the date of her final moult, and on May 13, but he avoided 1903.] NATURAL SCIENCES OF PHILADELPHIA. 123 her each time. On May 15 he completed his final moult, but on the next day he avoided her again; this would show that sexual ardor may not show itself in the male immediately after the final moult On May 17 they were put together at 8.00 P.M. Both remained perfectly quiet within an inch up to 8.57, when she moved toward him with her first two pairs of legs outstretched; he quickly reached out his corresponding legs, seized the feet of her right fore-legs with his right fore-legs, and holding them thus (only the metatarsi in contact) he quickly revolved around her and climbed upon her back from be- hind. She immediately became immobile with her legs flexed close to her body. He turned about upon her back, holding her with his legs, and feehng her with his palpi. At 9.02 he reached his head downward over the posterior end of her abdomen, placed the ventral surface of his cephalothorax against the ventral surface of her abdomen so that his head was posterior to her epigynum, his abdomen accordingly extending behind hers and tilted upward; his third and fourth pairs of legs held tightly to her abdomen, while his first and second pairs were directed forward to embrace her legs. One palpus at a time he then extended straight forward and applied to her epigynum. The number of palpal appUcations was, using the abbreviations r and I for right and left respectively, the first numbers expressing the time extent, the second the number of insertions: r, 9.08-9.12, 8; 1, 9.12- 9.13^, 6; r, 9.13^-9.36^, 9, the last time inserted continuously from 9.21 to 9.36^. The black spine connected with the palpal organ was pressed into her epigynal aperture, and when so inserted an evagi- nated sac connected with the concave surface of the organ expanded and collapsed after the withdrawal of the palpus from the epigynum; before another insertion the palpal organ was drawn through the chelicera. At 9.38 he moved half an inch away, and she rose suddenly at 9.41. The same 9 , No. 4, copulated several times with this male. No. 5 On each occasion their attitudes were the same. The following notes give concisely the duration of time and the number of applications of the palpi, as well the mode of approach of the male. (1) May 18, as soon as he touched her he immediately mounted upon her back, at 12.09 P.M.; palpi applied as follows: r, 12.16^-12.21, 9; r, 12.21-12.31^, inserted continuously. Then he walked half an inch away, returned and copulated again; r, 12.37-12.41, 2 or 3; he then left and again tried to return, but she repulsed him. (2) May 19, by touching him I brought him in contact with her he mounted her, and made the following palpal applications : 1, 8.28- 124 PROCEEDINGS OF THE ACADEMY OF [Jan., 8.30, 4, not effectively inserted; 1, 8.31-8.33, 1; 1, 8.34-8.34^, 1, 1, 8.35^8.36; 1, 8.36^8.42; 1, 8.421-8.43; 1, 8.43^-8.44; 1, 8.44-8.46 1; 1, 8.46^-8.47; 1, 8.47^-8.50; 1, 8.50-8.51; 1, 8.53-8.53*; 1, 8.54^- 8.57^. Then she rose and shook him off. (3) May 21, at 10.17 P.M., he touched her, kept quiet for half a min- ute, then climbed upon her. He left her for a few seconds at 10.25 then returned and applied his palpi as follows: 1, 10.29^-10.35, 5; r, 10.35^ to 10.36, 6; 1, 10.38^ to 10.41; r, 10.44^-10.46, several times ; 1, 10.46 to 10.48, several times; 1, 10.48-10.48+, 1; r, 10.48^-10.49^ 1; r, 10.49^-10.51; r, 10.51-10.52^ 4; 1, 10.53-10.55, 4; 1, 10.55-10.58, 1 ; r, 10.58-11.08, continuously; 1, 11.08-11.12, 1; r, 11.12+-11.17. He then got upon her back, remained quiet there for a while, then I sepa- rated them. (4) May 23, he mounted her at 3.34 P.M. Then the two being in an unfavorable place for study, I pushed them to another place with a pen-handle; she gripped it, and I was obliged to shake roughly to make her leave go, but he did not loosen his hold upon her, and as soon as she had lost her grasp upon the pen-handle she immediately became motionless again with flexed legs. He applied his palpi as follows: r, 3.39-3.40, 6; 1, 3.42-3.45, 1; r, 3.45*-3.46, 1; 1, 3.46-3.46^ 1; 1, 3.47-3.50, 1; 1, 3.51-3.52, 1; r, 3.521-3.54; 1, 3.54-4.00; then he walked off for a minute but returned; r, 4.03-4.04^, 1; 1, 4.04^-4.07; r, 4.08-4.11 ; 1, 4.12-4.13. He then left her, returned, did so again, but at 4.14 she arose. (5) May 28, he was placed with her from 9.24-9.50 P.M., but there was no copulation. (6) May 29, the male mounted her at 8.45 P.M. Palpal appli- cations: 1, 9.53-9.561 6; r, 9.56^-9.58*, 2. Then she began to walk about and carried him out of the open cage, and in putting them back I separated them. (7) June 3, 8.31 P.M. She moved so as to touch him, and a minute later he mounted her. Palpal applications: 1, 8.37*-8.39, several times; r, 8.39-8.40^, several times; 1, 8.40^8.45, continuously; 1, 8.45*-8.57, continuously; r, 8.57*-9.01, several times; r, 9.01-9.03. At 9.05 he got upon her dorsal side again, a minute later she rose and walked off with him chnging to her. She dislodged him at 9.09, but a minute later he mounted her again, used his right palpus from 9.12 -9.13, then she rose and escaped from him. $No. 109 copulated with c? No. 108 on May 24, and on May 25; the attitudes were as in the previous cases. On May 29 she was rather hostile to him, and whenever he tried to grasp her she ran off. Once 1903.] NATURAL SCIENCES OF PHILADELPHIA. 125 with his first two pairs of legs he parried with hers for more than a minute, evidently trying to grasp them; then he sidled off. and when he returned she grasped him with her legs, and finally she gripped fiercely at him. The descriptions here given may seem unnecessarily full, but they are given to show the variations in the mode of copulation, a greater variation than I have observed in any other species. In one copula- tion only one of the palpi was used; in the others both palpi, but in irregular alternation. The earlier palpal applications in each copula- tion are mainly ineffective, the palpal organ not being thoroughly in- serted, and a considerable part of the time is occupied in working the palpal organ through the chelicera. Peculiar for this species is the fact of the male sometimes interrupting the copulation by leaving the female for a minute, then returning and resuming it. The longest period of copulation was 4S minutes. There is no coiu'tship on the part of the male, nor any instigation by the female. When the male is first put into the cage of a female, even though he faces her, he does not seem to recognize her as such by sight, and pays no attention to her until he touches her, when he quickly gets upon her back. My custom was to push him up to the female. He is considerably the smaller and the more active; and when upon her back he is secure from her legs, and is not easily dis- lodged by her. He mounts upon her from behind, out of reach of her long and dangerous fore-legs, and when he is mounted she immediately becomes still. His smaller size and greater activity enable him to accomplish his purpose, quickness overcoming strength. Seeing the male upon the female, one involuntarily thinks of the Old Man of the Sea, who cannot be dislodged. Cocooning. — For some two weeks before making her cocoon 9 No. 4 spun a network of web-lines across her cage, so dense that she frequently became entangled in them, and I was obliged to clean them away. But this is probably not to be considered a web-making, but simply the habit this species has of drawing out a thread behind when walking, a "drop-line" by which it can find its way back. On July 5, however, she made quite a dense sheet of silk in a corner of the cage, completely enclosing her and inclined at an angle from the floor to the walls. Within this she made, in the early morning of July 9, her only cocoon; this cocoon was inclined at an angle to the floor, fastened to the scaffold- ing, flattened, unevenly circular in outline, and white in color. It did not hatch. 9 No. 109 built on July 5, at the angle of two walls of her cage with 126 PROCEEDINGS OF THE ACADEMY OF [Jan., the floor, a similar sheet-like tent; and on July 9 made her cocoon within it. This cocoon was roughly quadrangular in outline and flattened. It did not hatch. Care of the Young. — The cocoon is guarded by the mother, who lies upon it and embraces it tightly with her legs. 9 No. 4 would not leave her cocoon to secure living food during the first ten da3^s, but clung tenaciously to it ; but after that she would leave it to chase prey, and afterward would return to the cocoon, or at least to the silk sheet placed before it. By chmbing many times upon this silk sheet she gradually demolished it. 9 No. 109 at first held her cocoon zealously, then neglected it for several days, and finally at this time, October 4, is embracing it again. Philodromus aureolus Walck. Cocooning.—A female of this species made 6 cocoons, on she follow- ing dates respectively: July 3, 30, August 9, 21, 27, September 6. All these were normal cocoons except one, whicli I disturbed in the pro- cess of making, but the eggs of none of them hatched. The first three cocoons were placed side by side at the angle of the wall and the roof of the cage, the others at the angle of the wall and the floor. The mother constantly stands with outstretched legs over the most recent cocoon. Each cocoon is a flattened disk of rather loose white threads, and the mode of architecture shows that the mass of ova had been laid upon the surface of the lower disk (base), and then overspun with another disk (cover). But a peculiarity is a third disk of silk, concentric with but at a distance of about 2 mm. from the cover of the cocoon — the final protection, on which the mother stands. Prostliesima sp.s Cocooning. — On June 14, at 8.45 P.M., I saw a female just finishing her cocoon, which was discoidal in shape. She placed her feet upon its margin, and brushed her spinnerets from side to side across its surface. This was her second cocoon, and at 10.05 she started spin- ning again upon the wall of the cage, used her spinnerets as before, and continuing steadily up to 11.50, when I ceased to watch her. She had then made a thick silken covering, about four times her length in one direction, and half so long in the other. But she did not oviposit upon this surface, so that I cannot tell whether it was an abnormal cocoon or not. 'Mr. Banks wrote me that this species comes closest to P. insularis Ban-ks, but is different from it. 1903.] NATURAL SCIENCES OF PHILADELPHIA, 127 Thargalia brivittata (Keys.). Cocooning. — The end of this process only was seen. The discoidal cocoon was satiny-white and placed at the angle of the floor and wall of the cage ; its base was closely apposed to these surfaces so as to be bent at a right angle, while the cover (of smaller diameter than the base) was arched from the floor to the wall. In completing this cover the spider brushes her spinnerets more forward and backward than from side to side, flexing her body about energetically, while keeping her feet in one position upon the edge of the cocoon; then she changes her position over the cocoon, and repeats this process. I watched her in this process for half an hour, after which she covered the surface of the cocoon with small particles of dirt. Drassus neglectus Keys. Guarding of the Cocoon. — A female of this species was caught on June 10, and made a thick web within her cage. On the evening of July 7 she made her cocoon, which was white in color, discoidal with circular outline, one side flattened and the opposite side somewhat arched. This cocoon was loose, not fastened to any object, and she held its margin with her chelicera, pressing her cephalothorax against it and at times partially embracing it with her legs. Most of the time she held it in one corner of the cage, but sometimes carried it about to different portions of the cage. When I touched her she still clung to it without moving ; flies were put in her cage almost daily, but she paid no attention to them, even though at times they touched her. She died on July 21, still holding the cocoon in death. This observation is narrated, since I know of no other Drassid that guards its cocoon so carefully, nor of any that makes such a thick and large web. General Considerations. History of Our Knowledge. — In 1701 Leeuwenhoek, the father of histology, wrote: "I never was so happy as to see the Spiders couple, .... but what shall we say, the Coition of Spiders must differ funda- mentally from other Creatures, since their Matrix is placed in the vipper part of their Belly." He did not know of the observations of Lister, who in 1678 discovered that the male spider fertilizes the female by applying the enlarged terminal joint of his maxillary palpus to her epigynum. The discovery has been confirmed by a long list of obser- vers, though a list small in comparison with the number who have sought in vain to see the process : Ausserer, Bertkau, Blackwall, Camp- bell, Clerck, De Geer, Duges, Emerton, Fickert, Hasselt, Hentz, Her- 128 PROCEEDINGS OF THE ACADEMY OF [Jan., man, Lendl, Lesser, McCook, Menge (in numerous species), the Peck- hams, Ritsema, Seidel, Treat, Walckenaer, Westberg. These confirm- atory observations, embracing species of all the larger araneid families, make it very probable that in all modern Araneids the male transfers the sperm to the female by means of his palpi. Two writers have com- bated this conclusion: Treviranus, who first discovered the testes of spiders and, finding no organic connection between these organs and the palpi, maintained that the application of the male palpi to the epigynum is probably not real copulation, but rather a preliminary act of stimulation; and Cambridge, who states he observed two indi- viduals of a Lycosa in coition, with their genital apertures in apposition. In \dew of the numerous observations on this act in Lycosa, I have no hesitation in stamping Camljridge's assertion as erroneous. In all the families where the act has been seen, accordingly, the palpal organs are the transmitters of the semen. Now though Treviranus showed conclusively that there is no organic connection between the testes, or the vasa deferentia, and the peculi- arly modified palpal organs, some later writers, particularly Cambridge and Herman, insisted that there must be some such tubular connection, in order to explain the presence of spermatozoa in the palpal organs. The results of all other anatomists, however, have corroborated Treviranus, and Menge discovered, first in the year 1843, how the sperm is brought from the genital aperture into the tubular appa- ratus of the palpal organs. In that paper of 1843, an arachnological classic, he described the process for Linyphia triangularis and Agalena labyrinthica, showing that the male constructs a little silken bridge or "Steg," deposits a drop of sperm from his genital aperture upon it, then applies his palpal organs alternately to the drop until they have absorbed it all; then, and not till then, is the male ready for copulation. This process of charging the palpi with sperm, which I have termed here "sperm-induction," was described by Menge later also iovTapi- nopa longidens, Agalena similis and Micrommata virescens. The only other writers who have described this process are Black wall (1863, in Agalena labyrinthica), Ausserer (1867, in Dictyna henigna and Liny- phia triangularis), Bertkau (1875, 1876, in Philoica domestica, Linyphia montana and Cluhiona comta), Westberg (1900, in Linyphia triangu- laris). To these may be added my present observation on Lycosa stonei, L. ocreata pulchra, Tegenaria derhami, Theridium tepidariorum and Dictyna volupis (here seen twice). Hasselt was unable to see this process himself, and on that account doubted whether it is of general occurrence among spiders. 1903.] NATURAL SCIEN'CES OF PHILADELPHIA. 129 The Mode of Embrace in Copulation. — Grouping together my own observations with those of other natiirahsts. and bracketing the author- ity for each case, the following modes of embrace may be distinguished in spiders: (1) — The male above the female with his ventral surface apposed to her dorsal, their heads pointing in opposite directions, (a) — Female with her sternum down: Lycosa stonei, L. ocre'ata pulchra, L. scutulata (Montgomery), L. monticola (Clerck), L. amentata, L. rurestris (]\Ienge); Trochosa infernalis (Lendl), T. terricola (Menge); Pardosa yiigr opal pis (Mont- gomery) ; Sparassus (Bertkau) ; Epiblemum scenicum, Attidcc, Enophrys reticulata (Menge), Attus scenicus (De Geer) ; Plwurolithus festivus, MelanopJiora nocturna (Menge). (b) — Female partially upon her side: Agalena ncevia (Emerton, Montgomery), A. labyrinthica (Duges, Menge, Walck- enaer), A. similis (Menge). (2) — The head of the male facing the head of the female, their bodies in one line, (a) — ^lale with his dorsum directed upward, female partially upon her side: Tegenaria derhami (Montgomery). (6) — Both with ventral surfaces directed upward, bodies hori- zontal: Linyphia marginata (Emerton, ]\IcCook), ?Stea- toda boreal is (Emerton). (3) — Ventral surfaces of the pair apposed, sternum against sternum, heads directed in opposite directions, (a) — Copulation upon a web: Argenna (Bertkau), Linyphia mon- tana (Clerck), Linyphia triangularis (De Geer, Menge). (b) — Copulation upon the ground, (&') — Male below the female: Argyroneta aquatica (Walck- enaer), Chiracanthium oncognathus (Menge). {b") — Male above the female : Clubiona triviolis (Menge). (4)- — Male holding to the side of the abdomen of the female, his long axis at right angles to hers: Acrosoma gracile (Montgomery). (5) — Male with his ventral surface apposed to the posterior and ventral surface of the abdomen of the female, their heads turned in the same direction: Xysticus audax (Prach), A', triguttatus (Emerton), X. stomachosus (Montgomery), Misumena vatia (De Geer), Micrommata virescens (Menge). (6) — Ventral surface of the male turned toward the ventral surface of the female, their heads tiu*ned in the same direction, copula- tion in a web, (a) — Bodies not in actual contact, but heads converging: Epeira labyrinthea (^Montgomery), E. diademata (Menge, Lendl), Theridium tepidariorum, Pholcus phalangioides (IMont- gomery). 9 130 PROCEEDINGS OF THE ACADEMY OF [Jan., (6) — Long axes of the bodies converging so that the forehead of the male touches the sternum of the female: Dictyna volupis (Emerton, Montgomery), D. henignum (Walck- enaer), Micryphantes rurestris (Hasselt). (c) — ]\Iale with his sternum apposed to that of the female: Dictyna armnophila (Menge). (d) — Male apposed to the ventral surface of the abdomen of the female: Nephila, Argiope (Emerton), Meta segmentata, Asagena serratipes (Menge). (e) — ]\Iale holding the chelicera of the female with his own, the abdomen of the female flexed ventrad : Pachygnatha listeri (Menge), Tetragnatha extensa (Bertkau. Duges, Menge, Lister, Walckenaer). The following table illustrates the distribution of these modes of embrace among the families observed : Family. LycosidcB . ... SparassidcB . AttidcB DrassidcB . . . AgalenidcB . . T heridiidoB . . Argyronetidce EpeiridcB . . . Thomisidm. . . PholcidcB DietynidoB . . . Mode of Embr.^.ce. la 16 2o 26 3a 36 i 4 5 Ga i 06 t 6c 6d 1 6e + + + + + + + + + The species which have been observed are too few to allow broad generahzations, but the following points may be noted. In all the species it is the male that embraces the female, she remaining motion- less ; the male embraces with his legs, and in Tetragnatha and Pachyg- natha with his chelicera also. The mode of embrace may well be dependent upon one of several different factors: (1) relative size of male and female; (2) differences in their form; (3) differences in place of copulation (whether upon a web or upon the ground). The fact that the typical terrestrial spiders which do not build webs (Lycosiclce, Sparassidce, Attidoe, certain Drassidce) all show the same mode of embrace might point to similarity of place of copulation, in unison with the running and hunting habits of these species, inducing similarity of mode of embrace, and be referable to the third factor; 1903.] NATURAL SCIENCES OF PHILADELPHIA. 131 while the diversity of embrace in the web-making species may be in part due to differences in the web construction, in part to degree of sexual dimorphism. The Epeiridcc and Theridiidce show the greatest diversity of embrace, and in these families is to be found also the great- est sexual dimorphism. In other words, the mode of embrace may be brought into more or less conformity with the general mode of life of the species. Similarity of embrace points, then, to similarity of other habits, but not necessarily to community of descent. No spider shows the attitude of embrace common among insects, namely the male above the female with his head pointed in the same direction as hers. Details of the Process of Copulation. — The following table combines data of the other observers with my own upon the details of copula- tion. Where a species has been entered more than once, each repeti- tion denotes a separate act of copulation. In the column "Duration of Copulation," h, m, s, signify hour, minute and second respectively. In the column headed ''Simultaneous," the sign X denotes that both palpi are inserted at once, and the sign — that they are not. In the column headed "Alternated," the sign X indicates that the right and left palpi are inserted alternately, but that the number of alternations was not noted ; the sign — ,that only one palpus is employed ; while a number expresses the number of alternations of the right and left palpi. In the columns headed " Right Palp" and " Left Palp," a num- ber enclosed in brackets denotes the average number of times one pal- pus is inserted successively before the other palpus is inserted, this average being obtained by dividing the total number of in- sertions of that particular palpus by the total number of alternations of the palpi. Spaces left vacant in the columns denote lack of observations. For the great number of small details that cannot be reproduced in such a table, the reader is referred to the obser- vational part of this paper, and to the works mentioned in the literature list. 132 PROCEEDINGS OF THE ACADEMY OF [Jan., d, C- ^^_^-r^ 00 ^^ ^_, oj 00 Pm •^«o"^*- io o CO 1—1 r3 •* lO CO CO O 00 00 U-t Ph O t-i M a 'C s 0) ^ a g »0 CO -* O 00 l- ■^ C5 T-H 1—) ^ 0(N 50-* OS «o tH 1— 1 1— 1 T-H z^ C3 O l—t c3 cS l-> o Ol fH > 03 3P-1 ■+ I + ++ f I S I o «i CO I > o S ' g| I I + I +1 i I a 3 LO *>• ^ ^" S S ^ ^ ^ £- OS t- CO CO 00 i ^ . - ^- > ,: 10 M ^- a o a o <» <» g g g V "^ -^ K^ K^ k.( M, ., „ ., a S"? a *"• w »-< o n (ij S-* C^ >y. ^ S O r- ni C3 (-15 r- ;^' ^ ^ Qj -» 1,1; ^ s^ ^-^ g c^ '—' O OJ CB CJ . '^ . °^ ■otx tc: cr ~ie ts ??5 ~-^ kq;^K^^i;^^;^l^^^^^^ft,;^f>l^^■ § e s i= '^5 2 ^ ^ !S 5^ I^ ■T| >il &5 ^ E^ fe] k;^^c: s « S 2 .53 >5 •a^aisooA^ •a'aiaiatig; •a'aiiaiTianx 1903.] NATURAL SCIENCES OF PHILADELPHIA. 133 f -i rH i-H o d o LO s r-l X! a 3 ;zi a; a CO S Y-H 1—1 ia-^r-<-r-lr-ly-tCiOiT-lT-t -rHi-HCO eO;3 1 ^ 5 § 3 OJ o ;-i H o §..• IS + ++ 1 1 CO+, 1 1 1^^ 1 1 1 0^ CI T-H l-H T-l l-H tH T-H r a. OS "3 ^1 So^ , M 1* + MI 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 M 1 1 1 1 ^ • 3 2 2 -d CO rd ^- M 00 3 CO " j3 B.2 "^o -^ ^^ 'S -^1. S3 i a d^ o in *> ec « tn ^ c« ^ !S e « 2 §= ~ ■» ~. ■" ^ ~ S 'o m SS a- is ■=> s 2 « O (S Jt Si. S e °° e s ■3;S l^.o: 3 00 ft ^•8 s « K Iff § Si §-?§! g^S §-§ s S S « g Itll .S 'S l50 <^ f^ "^ i«;i^^Si'^i^ qqcJS^i^^^t^^S ^;^^^^t«;«^-;2 ^Oh^^ •a^ai laiua HX •a a im: I2t l:J IC I Ft ima T\ w V TFaissv" aa 134 PROCEEDINGS OF THE ACADEMY OF [Jan., aj s ji, "3 03 Pm 1 ^ ^ a is S -I-H T-H O 1-H 1-H tH a CO o '■6 # CO t- (M (M 0» ^ C-i 1-1 tH CO I-H "3 ■«^ o H (h'C G^ '■G Ci< S2 3g 1 _L ICO 1 1 «0 cr lO III ■2|£ 1-= 1 i 1 1 1 1 1 1 1 + + + CO o3 a «4H > ui . C.2 .2;! • "Ss _S oJ g ■1^ -fci o "^ C^lCOOOO-^ift OJkS'3 Iff c* oj oj -^ CO w co^-*q CO >% ;-i . r^ ^ ■s o o 3 1^ s s s ■< ««0*'®'»'S)'^ <©<»<» ^ ;?(^^^;^;^:?s as Sixth English Edition. Philadelphia.. *1878. Lebert, H. Bail und Leben der Spinnen, Berlin. 148 PROCEEDINGS OF THE ACADEMY OF [JaFx., 1701. Leeuwexhoek. A Letter from Mr. Anthony van Leeuwenhoek, F.R.S. concerning Spiders, their Way of KiUing their Prey, Spinning their Webs, Generation, etc. Phil. Trans. Roy. Soc. London, 22. 1886. Lendl, a. Ueber die Begattung der gekronten Kreuzspinne (Epeira diadeniata CI.). Term. Fiizetek, Budapest, 10. 1887. Lendl. Ueber die Begattung und die Copulationsorgane von Trocliosa infernaUs Motsch. Ibid., 11. 1742. Lesser. Theologie des InsCctes. Traduit .... par Mr. P. Lyonnet. T. 1. LaHaye. 1748. LiGNAC, J. A. L. de. Mcmoire povr servir a commencer Vhistoire des Araignees aquatiques. Paris. 1678. Lister, M. Historice Animalium Angliu; . . . . De araneis. London. *1778. Lister. Naiurgeschichte der Sjnnnen iibcrhaupt , etc. Quedhnburg and Blankenburg. 1871. Lucas, H. (De I'Ohos venatorius.) Ann. Soc. Entom. France Bull. de Seances. 1829. Lyonnet. Anatomic de differentes Especes d'Insectes. Araignees. Mem. Mus. Hist. Nat. Paris, 18. 1879. McCooK, H. C. Pairing of Spiders, Linyphia marginata. Proc. Acad. Nat. Sci. Philadelphia. 1884. McCooK. How Lycosa Fabricates Her Round Cocoon. Ibid. 1890. McCooK. American Spiders and Their Spiiining Work. \o\. 2. Pliihi- delphia. *1793. Martyn, T. Aranei: or the Natural History of Spiders, etc. London. 1843. Menge, a. Ueber die Lebensweise der Arachniden. Neueste Schr. naturf. Ges. Danzig, 4 1866. Menge. Preussische Spinnen. 1. Schr. naturf. Ges. Danzig (N.F.), 1. 1868. Menge. Idem, 3. Ibid., 2 1871. Menge. Idem, 4. Ibid. 1872. Menge. Idem, 5. Ibid, 3. 1873. Menge. Idem, 6. Ibid. 1874. Menge. Idem, 7. Ibid. 1876. Menge. Idem, 9. Ibid., 4. 1877. Menge. Idem, 10. Ibid. 1878. Menge. Idetn, 11. Ibid. *1790. Meyer, F. A. A. Ueber einige Spinnen der Guttinger Gegend. Gottingen. *1866. Ohlert, E. Ueber das Leben der Spinnen. Altpreuss. Monatschr., 3. 1887. Peckham, G. W. and E. G. Some observations on the Mental Powers of Spiders. Journ. Mar ph., 1. 1889a. Peckham. Observations on Sexual Selection in Spiders of the Family Attidse. Occas. Papers. Nat. Hist. Soc. Wisconsin, 1. 18896. Peckham. Protective Resemblances in Spiders. Ibid. 1890. Peckhaji. Additional Observations on Sexual Selection in Spiders of the Family Attidic. Ibid., 2 1899. PoKRowsKY, S. Beobachtungen iiber das Eierablegen bei Pholcus. Zool. Anz., 22. 1866. Prach. Monographic der Thomisiden (Krabbenspinnen) der Gegend von Prag, etc. Verh. zool.-bot. Ges. Wien, 10. 1798. Quatremere-Disjonval. Araneologie odcr Naturgeschichte der Spinnen, etc. Aus. d. Franzosischen d. 2ten Ausgal^e iibersetzt. Frankfurt a. M. *1872. RiTSEMA, C. C. Waarneming omtrent de copulatie der Spinnen. Tijdschr. v. Entom., 15. 1847. Seidel. Einige Beobachtungen an Spinnen. UebersicJtt d.. Arbeit, u. Verand. d. Schles. Ges. raterldnd. Kidtur, Breslau. 1864. Simon, E. Ilistoire naturelle des Araiqnies. Paris. 1874. Simon. Les Arachnides de France. Paris. *1882. SoRENSEN, W. Sur le rapprochement des sexes chez cjuclques Araignees. Tidjschr. v. Entom., 1. 1866. Termeyer, R. M. de. Researches and Experiments upon Silk from Spiders, and upon their Reproduction. Transl. from the Italian by B. G. Wilder. Proc. Essex Inst., .5. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 149 1870-1873. Thorell, T. Remarks on Synompns of European Spiders. Upsala. *1849. ToTi, L. Ueber die Lebensweise der Spinnen. Neujahrsstiicke Zurch. naturf. Ges. 1879. Treat, M. The Habits of a Tarantula. Amer. Nat., 13. 1812. Treviranus, G. R. Ueber den inneren Ban der Arachniden. Ntirnberg. 1888. Wagner, AV. La Mue des Araignees. Ann. Set. Nat. (7) Zool., 6. 1894. Wagner. LTndustrie des Araneina. Mern.Iniper.Acad.Sci.St.Peters- bourg (8), 42. 1837. Walckenaer. Histoire naturelle des Insectes, Apteres. Suites a Buffon, 2. Paris. 1891. Warburton, C. The Oviposition and Cocoon-weaving of Agelena labyrinthica. Ann. Mag. Nat. Hist. (6), 8. *1893. " Warburton. On the Oviposition of Agelena labyrinthica. Proc. Cambridge Sac, 7. 1900. Westberg, p. Aus dem Leben der Spinnen, 5. Zur Begattung von Linj'phia triangularis (var. macrognatha und micrognatha Menge). Kor- respendenzbl. naturf .-Ver. Riga, 43. *1828. Westwood, J. O. A Note upon tlie Impregnation of the Arachnida. Zool. .Journ. London, 4. ___„ Explanation of Plates IV and V. All the drawings are freehand outline sketches, and as such do not claim to be accurate in all dimensional proportions, though the attitudes are true to nature ami all are studies from life. The degree of magnifi- cation varies in the different figures. Plate IV, Fig. 1. — Lateral view of a female Ocyaleundata, hanging in her web and holding her cocoon. Fig. 2. — Copulation of Theridium tepidariorum , lateral view at the instant when the male seizes the female. Fig. 3.— Copulation of Pholcus phalangioides, only the proximal portions of the legs shown. Fig. 4. — Copulation of Agalena noivia, the female below and partially upon her side, the male's left palpus applied to the epigynum. Fig. 5. — Copulation of Acrosonia gracile, the male with his left palpus ap- plied to the epigynum ; the male is drawn proportionately some- what too large. Pl.\te V, Fig. 6. — Copulation of Dictyna volupis, the left palpus of the male inserted; the third leg of the male and the posterior end of the abdomen of the female are supported by the web. Fig. 7. — Copulation of Xysticus stomachosus. Fig. 8. — Copulation of Pardosa nigropalpis, lateral view, the male with his left palpus inserted. Fig. 9. — A female Pardosa nigropalpis holding her cocoon beneath her cephalothorax; this is the attitude when she finishes spinning the cover of the cocoon, and when she bites it open to free the young. Fig. 10. — Copulation of Lycosa stonei, dorsal view, the male inserting his right palpus, the cephalothorax of the female hidden by the male. 150 proceedings of the academy of [feb., February 3. The President, Samuel G. Dixox. M.D.. in the Chair. Fourteen persons present Papers under the following titles were presented for publication: "New Lithobii from California and Oregon," by Ralph V. Chani- berlin. "Descriptions of Several Fishes from Zanzibar Island, Two of which are New" by Henry W. Fowler. The death of Anthonv J. Antelo. a member, was announced. February 10. The President, Samuel G. Dixox, M.D.. in the Chair. Sixteen persons present. Papers under the following titles were presented for publication: "A Revision of the Orthopterous Genus Homoeogamia," by James A. G. Rehn. "Mollusca of Western Kansas and Adjacent States, with a Revision of Paravitrea," bj^ Henry A. Pilsbry. "A Revision of the North American Ants of the Genus Leptothorax Mayr," by William Morton Wheeler. February 17. Mr. Arthur Erwix Browx, A'ice-President, in the chair. Nine persons present. A paper entitled "Notes on the Conrad Collection of Yicksburg Fossils, with Descriptions of New Species," by Thomas L. Casey, was presented for publication. 1903.] natural sciences of philadelphia. 151 February 24. Mr. Akthur Erwin Brown, Vice-President, in the Chair. Twenty-eight persons present. Papers under the following titles were presented for pubUcation: "The Variations of Eutsenia in the Pacific Sub-region," by Arthur Erwin Brown. "A New Species of Nyctalemon," by Henry Skinner, M.D. Miss Lucy W. Baird was elected a member. The following were ordered to be printed : ^'^-^ PROCEEDINGS OF THE ACADEMY OF [Feb., NEW LITHOBII FEOM CALIFOKNIA AND OREGON. BY RALPH V. CHAMBERLIN. Since Dr. Anton Stuxberg published his Lithohioida Ajyiericce Borealis in 1875, very Httle has been added to our knowledge of the Lithobii of the Pacific coast region. A study of somewhat extensive collections made at various points in California and Oregon during 1902 enables me now to present descriptions of ten species not previ- ously recorded , as well as to identify with certainty various species described by earlier writers. In the key which follows I have included, in addition to the new forms, also such other related species as seemed proper for purposes of comparison: Analytical Key to Species Considered. A. — Coxal pores in several series (Bothropolys). Qj. — Angles of the 9th, 11th and 13th dorsal plates produced. 6i. — Articles of antennae 28-31 ; pairs of ocelli 30; coxal pores round or oval; length 15-18 mm., width 2.2 mm.; length of anal legs 6 mm., . aztecus Humb. et Sauss. bj. — Articles of antennae 42-46; pairs of ocelli 16-21; coxal pores strongly transverse (adults); length 22-23 mm., width 3.3-3.4 mm.; length of anal legs 10.5 mm., chumasanus sp. nov. a^. — Angles of none of the dorsal plates produced. bi.— Articles of antennae 20; pairs of ocelli 7-9 in 2 series; prosternal teeth 6-6; spines of anal legs 1, 4, 3, 1- 1, 4, 3, 2; length 18 mm monticola Stux. h^. — Articles of antennae 23-27 (normally); pairs of ocelh 17-21 in 4 or 5 series; prosternal teeth 8-8 to 13-13; spines of anal legs 1, 3, 2. 1 ; length 20-24 mm., I sierravagus sp. nov. B. — Coxal pores in a single series (Lithobius). flj. — Angles of the 9th, 11th and 13th dorsal plates produced. 6. J — Articles of antennae 26-28; prosternal teeth 2-2; pos- terior coxae unarmed laterally; spines of penult legs 1, 3, 3, 1, of anal 1 3, 2, 1 . . . . angelus sp. nov. 6,. — Articles of antennae 39-44; prosternal teeth 4-4; pos- terior coxae armed laterally; spines of penult legs 1, 3, 3, 2, of anal 1, 3, 3, 1, . . . hellulus sp. nov. (ij. — Angles of none of the dorsal plates produced. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 153 6j. — Posterior coxsp unarmed with spines; ocelli 6 in 2 series; porigerous area of coxse deeply depressed, pores .3, 4, 4, 3; anal legs of c? conspicuously compressed sublaterally, without produced lobes; length 8 mm., remex sp. nov. 6,. — Posterior coxae armed laterally (porigerous area of last coxse not deeply depressed). Cj. — Last 3 pairs of coxa? laterally armed. f/j. — OcelU 8-12; spines "of first legs 1,2, 1, of penult 1, 3, 3, 0-1, 3, 3, 1, of anal 1, 3, 3, 0; length 7.5-9 mm., eigenmanni Boll. f/„.— Ocein 12-15; spines of first legs 2, 3, 2, of penult 1, 3, 3, 2, of anal 1, 3, 2, 0; length 11.5-13 mm., . . . pitophilus sp. nov. Cj. — Last two pairs of coxse laterally armed. d^. — Anal legs in d^ not specially modified; claw of 9 gonopods tripartite; coxal pores 2, 2, 3, 3- 3, 3. 4, 3, very small; spines of first legs 1 , 2 , 1 , of anal 1 , 3 , 3 , 1 ; length 8-8.5 mm . , sastiaiius sp. nov. (h. — Anal legs in c? not specially modified ; claw of 9 gonopods tripartite; coxal pores 3, 4, 4, 3- 4, 5, 4, 4, large; spines of first legs 1, 3, 1, of anal 1, 3, 2, 0; length 9-10 mm., mesechinus sp. nov. r/3. — Tibia of anal legs in d^ produced above into a large peculiar lobe at distal end ; (9 unknown) ; coxal pores 2, 3, 3, 2, small; spines of first legs 1, 3, 2-2, 3, 2, of anal 1, 3, 2, 0; length 8.8 mm., castellopes sp. nov. d^. — Tibia of anal legs in c^ produced into a con- spicuous lobe at the superior-interior angle of proximal end ; claw of 9 gonopods tripartite ; coxal pores 3, 4, 4, 3, moderately large; ; spines of first legs 1, 3, 2, of anal 1, 3, 2, 0- 1, 3, 2, 1; length 10.5-11 mm., clavigerens sp. nov. f/5. — Tibia of anal legs in d^ excavated within near proximal joint, at which it and femur are produced inward into a conspicuous lobe; claw of 9 gonopods tripartite; coxal pores 2, 2, 2, 2,-2, 3, 3, 3, small; spines of first legs 1, 2, 1,-2, 3, 2, of anal 1,3, 2, 0; length 7.5-11 mm., utahensis Chaxnh. dg. — (c? not known.) Claw of 9 gonopods entire; coxal pores 2,3,3,3; spines of first legs 1 , 2, 1 ; of anal 1, 3, 2, 1 ; length 13 mm., ohesus Stux. 154 PROCEEDINGS OF THE ACADEMY OF [Feb., 1. Lithobius chumasanus sp. nov. Description. — Ackilts deep-brown to mahogany, legs paler distally, antenna paler distally or not; robust, head and dorsum roughened, more strongly so caudally; entire body rather sparsely pilose with very short hairs, but the last ventral plates more densely provided with longer hairs; antennae on basal joints very sparsely, elsewhere sub- densely, clothed with straight stiff hairs; legs sparsely hirsute, the last tarsal joint of anterior legs more densely so beneath. Head rounded, very nearly equal in length and width. Antennae moderate, article? 42-46, except the first ones short, the ultimate in length about equal ing the two preceding together. Ocelli on each side 16-21, arranged in 4 or 5 longitudinal series, large and distinct, the single ocellus ver;y large and vertically oval. Prosternal teeth 6-6, 7-7, large and sub- equal. Spines of first legs 2, 3, 3; penult legs with two claws, spines 1, 3, 3, 1 ; anal legs with two claws, spines 1, 3, 3, 1 ; last 3 pairs of coxae armed laterally and also dorsally with a stout spine. Coxal pores 7, 8, 8, 6-8, 8, 8, 7, transverse, oval in immature specimens. Gonopods of ? : claw stout, bipartite, the lobes equal or nearly so, sometimes with a very sifiall third lobe on the inner side of claw; basal spines, stout, pointed, flattened distally. Length of body 22-23 mm.; width of 8th dorsal plate 3.3-3.4 mm., length of antennae 10 mm. ; length of anal legs 10.5 mm. Habitat. — Santa Barbara, Cal. Etymology. — From Chumashans, a tribe of Indians of southern Cali- fornia, including those of the Santa Barbara mission. 2. Lithobius sierravagus sp. nov. Description. — Chestnut, legs and antennae lighter, the latter pale distally (specimens from Tiuckee, Cal.) to mahogany, with the legs and antennae dark, both legs and antennae rufous distally (specimens from Oregon City, Oreg.); head and dorsum much roughened, espe- cially so the posterior plates; dorsum glabrous or nearly so; venter very sparsely provided with hairs, the anal and genital segments sub- densely so; legs sparsely pilose; antennae uniformly clothed with fine, straight hairs. Head scarcely wider than long (14: 13.5), the portion back of eyes subquadrangular. Antennae: articles 23-27, rarely 20. except the first ones, short, the ultimate mostly a little shorter than the two preceding together, the number of articles usually in inverse proportion to length of antennae. Ocelli on each side : 16-21, in 4 curv- ing series, mostly 1+6, 5, 4, 3 or 14-5, 6, 3, 2. Prosternal teeth: 8-8 to 13-13. black and conical. Spines of the first legs 2, 3, 2; penult 1903.] NATURAL SCIENCES OF PHILADELPHIA. 155 legs with 3 claws, spines 1, 3, 3, 2; anal legs with the claw single, some- times with a short spine on the outer ventral surface of base of claw, spines 1, 3, 2. 1: last 2 pairs of coxae armed beneath and laterally. Coxal pores numerous, in 3-4 series, of varying sizes. Gonopods of ? : claws short and stout, tripartite, the lobes short and blunt or rounded, the lateral ones sometimes scarcely indicated; basal spines 3-3, moderate, conical. Length of body 20-24 mm.; width of 8th dorsal plate 2.4( 9 )-3.9 mm. (c^); length of antennae 10-12 mm. ( ?), 11-16 mm (d^); length of anal legs 10-12 mm. Habitat. — Truckee and Shasta Springs, Cal., and Oregon City and Portland, Oreg. Etymology. — Sieira from Sierra Nevada Mountains, the home of the species, and Latin vogvs, participle of vagari, to roam or wander over. 3. Lithobius angelus sp. nov. Description. — Reddish-brown, legs paler, antennae light distally; dorsal plates some^^•hat coarsely punctate, the posterior ones more strongly so ; dorsum glabrous ; venter mostly nearly glabrous, but the posterior plates and the genital and anal segments with more numer- ous hairs; legs sparsely pilose, antennae densely pilose. Head wider than long (7.5 :7) (& ) , or subequal in length and width ( 9 ) . Antennae : articles 26(cJ')-28 (9), articles short, the ultimate about equal in length to the two preceding together. Ocelli on each side: 12 in 3 longitudinal series (1-1-3. 4, 4). Prosternal teeth 2-2, acute. Spines of the first legs 1, 3, 1-1, 3, 2; penult legs with 2 claws, spines 1, 3, 3. 1 ; anal legs with 2 claws, spines 1, 3, 2, 1. Posterior coxae armed neither beneath nor laterally. Coxal pores 5, 5, 5 (C), 5, round or suboval. Oonopods of 9 : claw tripartite; basal spines 2-2, rather stout. Length of body 14 mm. (o^)-15 mm. (9); width of 8th dorsal plate 1.9 mm.(cr')-1.7 (9) ; length of antennae 4.5 mm.( 9 )-5 mm. (d^). Habitat. — Los Angeles county, Cal. Etymology. — Latin angelus, an angel, the reference being to the locality where found, Los Angeles, "place of angels." 4. Lithobius bellulus sp. nov. Description. — Body shining testaceous-brown of a reddish tint, head with lower portion of antennae and the posterior dorsal plates darker; legs paler, with a violaceous tinge, legs and antennae lighter colored distally; head and anterior dorsal plates weakly finely rough- ened, the posterior dorsal plates more strongly so: head and dorsal 156 PROCEEDINGS OF THE ACADEMY OF [Feb., plates with a icw scattered hairs, more numerous on posterior plates; venter likewise nearly glabrous, except the last plates; antennte uni- formly clothed with straight slender hairs of moderate length. Head but little wider than long (10:9). Antennse: articles 39-44, mostly very short, the ultimate about equal in length to the two preceding taken together. Ocelli on each side: 11-13, arranged in 3 series (1 +3, 4, 3-1 +4, 4, 4), somewhat distant from each other. Prosternal teeth: 4-4, the 2 inner on each side rather closer together than others. Spines of the first legs 1,3.1; penult legs with 2 claws, spines 1, 3, 3. 2 ; anal legs with 2 claws, spines 1, 3, 3, 1; last 3 pairs of coxse armed laterally with a stout spine, unarmed below or above. Coxal pores I'ound, moderately large, 5, 5, 5, 4-5, 6, 6, 4 (5). Gonopods of 9 : claw broad, tripartite, the middle lobe longest, slender and acute, the side lobes short; basal spines 2-2, the outer stouter than the inner and flattened distally. Length of body 13-16 nmi. ; width of Sth plate 1.6-2 nnn. ; length of antennse 6.3-8 mm.; length of anal legs 5.4-7 mm. Habitat. — Portland., Greg. Etymology. — Latin bellulus, of pleasing appearance, graceful. 5. Lithobius remex sp. nov. Description. — Body brown, legs paler; head smooth; dorsal plates nearly smooth, a little roughened by weak longitudinal folds; head with a few long hairs ; dorsum \\\ih scattered short hairs ; ventral plates smooth and glabrous ; legs sparsely pilose ; antennae rather sparsely clothed with moderately long straight hairs ; prosternum with a few hairs about margin ; claws of prehensorial feet with long hairs at base. Head cordate, slightly wider than long. Antennse: articles 20, be- yond the second short, except the ultimate Avhich is long, nearly equal- ing in length the three preceding together. Ocelli on each side: 6, arranged in 2 series (1+2, 3), deeply pigmented. Prosternal teeth 2-2. Spines of the first legs 1, 2, 1; anal legs with the claw single, spines 1, 3, 2, 0, in thecJ' conspicuously expanded in a subvertical direction, the plane of flattening more nearly horizontal distally, not produced into lobes or knots; posterior pairs of coxse without trace of spines. Coxal pores: pore-bearing groove deep, so that the pores are often visible with difficulty from below, 3, 4, 4, 3, moderate. Length of body 8 mm.; width of Sth dorsal plate 1.1 mm.; length of antennae 2.9 mm. ; length of anal legs 2.9 mm. Habitat. — Shasta Springs, Cal. Etymology. — Latin remex, an oarsman, the reference being to the flattened anal legs of the d^. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 157 6 Lithobius pitophilus sp. uov. Descriptions. — Brown or chestnut, legs and ventral surface paler, antenna^ rufous at tips; dorsal plates smooth, with a few shallow longitudinal furrows, glabrous; legs and prehensorial jaws sparsely pilose; antennte moderately densely pilose. Head subcordate, wider than long (7.25:6.75, nearly). Antennae with 20 articles, the ulti- mate mostly shorter than the two preceding together. Ocelli on each side: 12-13 in 3 series (1+3, 5, 3, 4). Prosternal teeth 2-2, acute. Spines of first legs 2. 3, 2; penult legs having the claw with 2 accessory spines, spines 1, 3, 3, 2; anal legs with the claw unarmed, spines 1, 3, 2, 0, femur in o ' enlarged distally and the tibial and tarsal joints inflated; posterior coxae unarmed beneath, the last 3 armed dorsally and laterally toward the dorsal surface. Coxal pores 4, 4, 5, 4-4, 5, 5, 5, small and round. Gonopods of 9 : claw weakly tripartite or sul)- entire; basal spines 2-2, stout. Length of body 11.5-13 mm. ; width of Sth dorsal plate 1.6-1.8 mm. ; length of antennae 3.5 mm. ; length of anal legs 4 mm. Habitat. — Truckee, Cal. Etymology. — Greek -tV^^i-. pine, and fiV-o^. fond of , loving, i.e., "pine loving," the specimens having been collected in the pine woods al)out Truckee. 7. Lithobius sastianus sp. nov. Description. — Testaceous-brown, posterior borders of principal dor- sal plates mostly darker; legs paler; antennae brown at base, yellow distally, head and dorsal plates hot strongly roughened, the plates with distinct longitudinal rugae ; venter a little roughened ; entire body sparsely pilose, the legs more densely so; antennae moderately densely clothed with long stiff hairs. Head a little wider than long (5.5-5 A-ery nearly). Antennae: articles 20, short excepting the first few. the ultimate a little longer than the two preceding together. Ocelli on each side : 10, arranged in 3 series (1-1-4, 3, 2), those of the upper row largest, those of the middle row intermediate. Prosternal teeth 2-2, small. Spines of the first legs 1. 2, 1 ; penult legs with 1 accessory claw, spines 1, 3, 3, 1 ; anal legs with the claw unarmed, spines 1, 3, 3, 1 ; the last 3 pairs of coxae armed with a spine dorsally, the last 2 also laterally. Coxal pores very small, round, 2, 2. 3, 3-3, 3, 4, 3. Gono- pods of 9 : claw tripartite, the lobes all short, not very acutely pointed ; basal spines 2-2, acuminate, the two outer longer than the inner ones.. Length of body 8 mm.( 9 )-8.5 mm. (o ) ; width of 8th dorsal plate 158 PROCEEDINGS OF THE ACADEMY OF [Feb., 1.2 mm.(c^)-1.4 mm. (9); length of autemioe 2.2 mm.($)-2.8 mm. (c?); length of anal legs 2.2 mm.( 9 )-2.8 mm. (d). Habitat. — Shasta Springs, Cal. Etymology. — The Sastians or Shastans were a tribe of Indians for- merly holding the country about Mt. Shasta, the locality in which the specimens were collected. 8. Lithobius mesechinus sp. nov. Description. — Testaceous-brown, the legs and antennre yellow, head and dorsal plates smooth or scarcely roughened, glabrous: ventral plates sparsely pilose, the posterior ones more densely so; legs very sparsely pilose; antennae with the ultimate articles sparsely provided with stiff hairs, elsewhere very nearly glabrous. Head wider than long (7:6.25). Antennae: articles 19-21, short, the ultimate shorter than the 2 preceding taken together. Ocelli on each side : 13, arranged in 3 series (1 +5, 4, 3), pale, Prosternal teeth 2-2, pale, small. Spines of the first legs 1, 3, 1; spines of penult 1, 3, 3, 1-1, 3, 3, 2; anal legs with the claw unarmed, spines 1, 3, 2, 0; last 2 coxse armed laterally toward dorsum and also dorsally ; the. median ventral spines of legs conspicuously longer than the others. Gonopods of ? : claw tripar- tite, basal spines 2-2, acute. Length of body 9-10 mm.; width of 8th dorsal plate 1.5 mm.: length of antennae 2-3.5 mm,; length of anal legs 3.5 mm. Habitat. — ^Meacham, Oreg. Etymology. — Greek' iJ.t^^ a hedgehog, spiny object, etc., so named in reference to the long median ventral spines of the legs. 9. Lithobius castellopes sp. nov. Description. — Brown, legs and antennae paler, yellowish; head and dorsal plates smooth; head sparsely pilose with moderately long hairs; dorsal plates very sparsely pilose, more especially the posterior ones, particularly about the margins; ventral plates, prosternum and prehen- sorial feet sparsely pilose, the genital segment nearly glabrous below : legs sparsely pilose ; antenna subdensely clothed with rather long stiff hairs. Head nearly equal in length and width. Antennae : articles 20, mostly short, the ultimate equal to or a httie longer than the two pre- ceding taken together. Ocelli on each side : 12, in 3 series (1 +5, 4, 2). Prosternal teeth 2-2, small. Spines of the first legs 1, 3, 2-2, 3, 2; penult legs having the claw armed with 2 accessory spines or claws. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 159 spines 1, 3, 3, 1; anal legs with the claw unarmed, spines 1, 3, 2, 0, in the d^ much modified, enlarged, the tibia above at the distal end pro- duced into a large knotty lobe, the lobe truncated posteriorly, sparsely pilose, bearing upon its posterior surface a spine which projects cau- dally; the last 2 pairs of coxse armed laterally, the last 3 dorsally. Coxal pores very small, 2, 3, 3, 2 Length of body 8.8 mm.; width of 8th dorsal plate 1.3 mm. ; length of antennae 3 mm. ; length of anal legs 3 mm. Habitat. — Shasta Springs, Cal. Etymology. — Latin castellum, a citadel or fortified place, and pes, foot, a name suggested by the appearance of the enlargement on the tibia of the c? anal legs. 10. Lithobius clavigerens sp. nov. Description. — Color of body, antennae and last pair of legs brown, other legs yellowish; head and dorsum smooth and glabrous, the head sparsely minutely pimctate; ventral plates and the prosternum with prehensorial feet smooth, sparsely pilose, the last 2 together with the genital and anal segments subdensely pilose ; antennae pilose, the outer surface of the proximal articles subglabrous. Head sub-round, about equal in length and width, narrowed anteriorly. Antennae: articles 20, the ultimate about equaling in length the 2 preceding taken to- gether. Ocelli: 10-14, pale and distinct, in 3 series (1+5, 5, 3). Pro- sternal teeth 2-2, moderately small. Spines of the first legs 1, 3, 2, penult legs with 1 accessory claw, spines 1, 3, 3, 1; anal legs with the claw unarmed, spines 1, 3, 2, 0. Anal legs in 9 scarcely more slender than in d" ; the exterior and interior superior margins of tibial and tarsal joints extended laterally, making the upper surface flat, the femur compressed in a different plane; the tibia densely punctate beneath, the tarsal joints less densely so; the first tarsal joint with a knob-like swelling at proximal end (absent incJ^); penult legs similarly modified to anal and to a scarcely smaller degree. In thed^ the anal legs are also swollen along or a httle below the superior margins, but the expanded portions are thicker and not so suggestive of flattening ; tibia marked above for its entire length by a ridge-like swelHng which at the proximal end is produced upward and inward into a conspicuous pilose lobe, tibia sparsely punctate beneath, the tarsal joints more densely so ; in the penult legs the tibia is somewhat similarly modified to that of the anal legs, but the lobe at the proximal end is missing not so strongly modified as in 9 . Coxal pores small and round, 3, 4, 4, 3. Gonopods of 2 : claw rather large, tripartite, lobes all acute, 160 PROCEEDINGS OF THE ACADEMY OF [Fcb., the middle one much longest; basal spines 2-2, moderate, subequal, or the inner ones a httle longer. Length of body 10.5-11 mm. ; width of 8th dorsal plate 1.4-1.5 mm. ; length of antennas and of anal legs each 4 mm. Habitat. — Pacific Grove, Cal. Etymology. — Latin clavis, a club, and gercns, participle of gerere, to bear; so named because the swelling on the dorsal surface of the tibia in the d^ anal legs suggests a club with the large end placed proxi- mally. 1903.] NATUKAL SCIENCES OF PHILADELPHIA. 161 DESCRIPTIONS OF SEVERAL FISHES FROM ZANZIBAR ISLAND, TWO OF WHICH ARE NEW. BY HENRY W. FOAVLER. A small collection of fishes has recently been forwardetl to the Academy from Zanzibar Island, East Africa, by Sir Charles Eliot. At present all of the examples are contained in the collections of the Academy of Natural Sciences of Philadelphia. PLATACID^. 1. Platax vespertilio (Bloch). 17SS. Chcctodon vef^pcrtilio Bloch, Ichtliyologie, II, pt. 6, p. 49, Plate 199, fig. 2; Je Fai trouve parmi les poissons qu'on m'a envoyes du Japon. 1866. Platax vespertilio Playfair, Fishes of Zanzibar, p. 64; Aden; Zanzibar [No descr.] Head 2f ; depth equal to length; D. VI, 35; A. Ill, 26; P. 18; V. I, 5; scales, in the lateral line to the base of the caudal 66, between the origin of the soft dorsal and the upper curve of the lateral line about 24, and between the latter and the origin of the soft anal 38; width of head 1| in its length; snout 2J; eye 3^; maxillary 3^; interorbital space 2f ; pectoral H; least depth of caudal peduncle 2^; ventral If in body. Body very deep, greatly compressed, more or less trenchant ab()\-o and below, and the greatest depth about midway in the length of body and between the origins of the soft dorsal and anal fins. Cau- dal peduncle short, deep and strongly compressed. Head short, very deep, greatly compressed, and its greatest width at the preopercle considerably wider than any other portion of the bod}-. The snout is rather long, steep, blunt and rounded above. Eye small. high in the head, and a little anterior. ]\louth small. Inroad and nearly horizontal when closed. Maxillaiy small, obliquely vertical, and hardly reaching beyond the anterior nostril. Teeth small, sharp- pointed, and in rather narrow bands in the jaws. A'omer and palatines smooth. Mandible rather broad and slightly protruding, so that the anterior profile of the body forms a bliuit angle at this point. Tongue not distinct from the floor of the mouth. Anterior nostrils circular, with the space between less than the space between the posterior pair, 11 162 PROCEEDINGS OF THE ACADEMY OF [Feb., and situated much nearer the front rim of the orbit than the tip of the snout. Interorbital space rather broad and convex. Gill-opening oblique, and carried forward below the front rim of the pupil. Gill-rakers short, and in moderate number. Pseudobranchiae absent. Gill-filaments very long and numerous. Peritoneum silvery. Scales small, finely ctenoid, and arranged in rather narrow vertical series. The greater basal portions of the vertical fins covered with small scales. Snout, interorbital space, occipital process, jaws and lower surface of the head naked, otherwise covered with small scales. No pectoral flap. Ventral connected at the base by two frena, the outer scaled, and the inner joining the innermost ray with the belly. Lateral line continuous, strongly arched anteriorly. Dorsal spines graduated from the first, which is very short, to the last, which is many times its length. The anterior rays of the soft dorsal and anal are very elongate and produced, those of the former nearly as deep as the body. They become graduated posteriorly to the last which are very short. The anal spines are similar to those of the dorsal, only much shorter. Caudal small, and rather deep. Pectoral short. Ventral very long, the outer rays three times the length of the inner, or half the depth of the body. Color in alcohol dark brown, the soft dorsal and anal blackish. Caudal blackish at base and sharply defined from the rest of the fin, which is pale straw color. Pectoral pale brown. Ventral blackish. Length 3^ inches. Two examples. The larger, which is described above, agrees with Bloch's figure perfectly in the broad vertical fins and the dark sharply defined color on the base of the caudal. The other has the dorsal, anal, and ventral very much more elongate, and the colors paler. These fins are blackish, and the outer edge of the ventral is whitish. No traces are evident of the two vertical bands shown on the anterior part of the body, like those in Ruppell's figure of Platax alhipundat us . BALISTIDiE. 2. Balistapus aculeatus (Linna?us). 1758. [Balistes] aculeatus Linnaeus, Syst. Nat., Ed. X, p. 328; India. 1866. Balistes aculeatus Giinther, Fishes of Zanzibar, p. 134; Zanzibar. [No descr.] 1870. Balistes aculeatus Giinther, Cat. Fish. Brit. Mus., VIII, p. 223; Isle of France; Island of Johanna (Coll. Dr. Kirk); Zanzibar (Coll. Lieut.-Col. Playfair); West Coast of Africa; Moluccas; Amboyna (Coll. M. Frank); China Seas ; Pacific (Coll. M. Haslar) ; Feejee Islands; Micronesia (Coll. M. Wright) ; Seychelles (Coll. Prof. E. P. Wright) ; Mauritius. 1899. Balistes aculeatus Jatzow and Lenz, Abhand. Senck. Naturf. Gesel., XXI, p. .530; S. Juan de Nova: Zanzibar (Coll. Dr. Voeltzkow). [No descr.]. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 163 Head 2f ; depth li; D. III-25; A. 21; scales about 42 to base of caudal; width of head 2 in its length; depth of head, measured verti- cally over posterior rim of orbit, greater than the length of the head by a little over half an eye-diameter; snout If; eye 3f ; interorbital space 3f ; first dorsal spine ly^; base of soft dorsal If; base of anal If; caudal 2; least depth of caudal peduncle 3i; pectoral 2f . Body elongate, somewhat elliptical and strongly compressed. The greatest depth about the middle of its length, or in the region of the base of the ventral spine. Caudal peduncle compressed and rather thick. Head long, angular, the upper profile more or less straight, at least in front, and the lower profile evenly convex. Snout long, convex above, very oblique, and the cheek compressed. Eye small, high, very posterior, and about two-fifths the length of the snout. Mouth small, and with thick fleshy lips. Jaws nearly equal, the teeth pro- truding. Teeth large, powerful, the edges sharp and notched. Nostrils small, very close together, lateral and in front of the eye above. Inter- orbital space, together with the space posterior to the eyes, convex. Gill-opening small, about two-thirds the interorbital space. Peritoneum black. Scales all very rough, those on the head and abdomen all more or less in oblique rows, and those on the trunk in nearly vertical series. Tubercles on the back between the dorsal fins enlarged, much larger than the three horizontal series of spines on the side of the caudal peduncle. A number of rather long spines between the ventral spine and the origin of the anal fin. About five enlarged bones above the base of the pectoral behind the gill-opening. Origin of the spinous dorsal about over the gill-opening, the first spine very large, robust and rough on the anterior surface, second spine much smaller and slender, and the third minute. The fin is depressable in a groove. Soft dorsal and anal opposite, and the rays all of nearly uniform height or about equal to the length of the eye. Caudal broad, rather deep, and the margin straight. Pectoral broad, rounded and short. Ventral spine rather long, freely movable and rough. Color in alcohol faded more or less uniform brown. Side with a deep brown band running from the gill-opening to the anterior and posterior rays of the dorsal, leaving a pale area at the middle of the base of this fin. Below, and running obliquely backward to the base of the anal, are first one broad dark band, and then three narrower. The dark band from the eye down to the base of the pectoral is distinct, 104 PROCEEDINGS OF THE ACADEMY OF [Feb., and the dark interorbital space with four narrow connecting lines is also very distinct. Three series of tubercles on the side of the caudal peduncle blackish. Membrane of the spinous dorsal blackish, the other fins plain pale brown. Very indistinct traces of the pale longitudinal lines on the soft dorsal and anal are still evident. Length 2^^ inches. One example. TETRAODONTID^. 3. Tetraodon aerostaticus (Jenyns). 1842. Tetrodon aerostaticus Jenyns, Zool. Voy. Beagle, Fish, pt. 4, p. 152. [Locality unknown.] 1866. Tetrodon lineatus Giinther, Fishes of Zanzibar, p. 131; Zanzibar. [Not of Linnseus.] 1870. Tetrodon stellatus Gunther, Cat. Fish. Brit. Mus., VIII, p. 294; Port Natal (Coll. M. Ayres) ; Zanzibar (Coll. Lieut.-Col. Playfair) ; Amboyna (Colls. M. Frank, Dr. Bleeker. Type of Tetrodon astrotaenia Bleeker) ; Japan; Feejee Islands. [Part, not var. n and /?.] Head 2^; depth (not inflated) about 2§; D. 10; A. 10; P. 19; width of head greater than its length by nearly an eye-diameter; snout 2 in the head; eye about 6; interorbital space If; length of dorsal 2^; anal 2f; caudal 2; pectoral 3; width of base of pectoral about 3f ; least depth of caudal peduncle 4. Body short, broad, and the belly capable of being greatly inflated, the greatest width about the bases of the pectorals. Caudal peduncle not very long, compressed. Head large, very broad, rounded or strongly conic above. Side of the head more or less constricted, and the cheek not very convex. Snout rather long, rather obtuse, and its upper profile a little concave. Ej^e rather small, about two-fifths the length of the snout, high, and about midway in the length of the head. Mouth rather small, broad, its width nearly twice the eye-diameter. Lips broad, fleshy, covering the greater portions of the teeth. Teeth large, very powerful, the cutting edges concave and the suture without deep groove or ridges on each side. Bifid tentacle conspicuous, about the last third of the space between the upper lip and the front margin of the eye. Inter- orbital space very broad, nearly flattened in front, and becoming gradually slightly convex posteriorlv. The internasal space is about one-half as wide as the interorbital. Gill-opening nearly as wide as the base of the pectoral, or about equal to the internasal space. Peritoneum silvery. With the exception of the lips, leases of the fins, and caudal peduncle, 1903.] NATURAL SCIENCES OF PHILADELPHIA. 165 the body is covered with rather large conspicuous spines, which are very rough to the touch. Those on the upper surface are short and sharp, and those on the belly elongate and somewhat clavate. Origin of the dorsal nearer the tip of the caudal than the posterior margin of the eye, and the fourth ray the longest. Anal entirely behind the dorsal, the fifth and sixth rays the longest, and, like the other fins, rounded. Caudal elongate, deep, and the margin rounded. Pectoral broad, the upper rays the longest, and situated directly behind the gill-opening. Color in alcohol dark brown, especially on the back, the belly hardly paler. The back, upper portion of the head, caudal peduncle, and cau- dal fin, marked with numerous small round blackish spots. The dorsal, anal and pectorals plain brown without markings. The flanks are marked with about six oblique blackish bands, the first few broad- est, and none of them extending on the middle of the belly. Vent blackish. Bases of the pectoral and anal with several dark spots. Length 4^^ inches. One example. 4. Tetraodon immaculatus (Schneider). 1801. [Tetrodon] Immaculatus Schneider, Syst. Ichth., p. 507. [After Lace- pede.] ' 1S66. Tetrodon immaculatus Gi'inther, Fishes of Zanzibar, p. 132; Aden; Zanzibar. 1870. [Tetrodon immaculatus] var. immacidata Giinther, Cat. Fish. Brit. Mus., VIII, pp. 291, 292; Port Natal (Coll. M. Ayres); Zanzibar (Coll. Lieut.-Col. Playfair); Mauritius (Coll. M. Cuming); Penang (Coll. Dr. Cantor); East-Indian archipelago; New Caledonia (Coll. M. Cuming); Australia. (Colls. M. Gould, Earl of Derby.) Head 2f ; depth (not inflated) about 2f ; D. 8; A. 9; P. 17; width of head 1^ in its length; snout 2\ in the head; eye 6^; interorbital space 2; length of dorsal 2\; anal 2\] caudal l\; pectoral 2| ; least depth of caudal peduncle 3. Body rather short, compressed, moderately broad, and the belly capable of great inflation, the greatest width about the bases of the pectorals. Caudal peduncle rather short, deep and compressed. Head moderately large, rather broad, rounded and little depressed above. Side of the head not very convex. Snout long, blunt, and with the upper profile straight. Eye small, a little nearer the gill- opening than the tip of the snout. Mouth small, terminal, oblique , with the mandible protruding, and its width about equal to one and one-half eye-diameters. Lips broad, fleshy, and almost entirely cov- 1 Based on " Le tetrodon sans-tache" Lac^pede, Hist. Nat. Poiss., I, 1798, pp. 475, 486, plate 24, fig. 1. [From Commerson; no locality.] 166 PROCEEDINGS OF THE ACADEMY OF [Feb., ering the teeth. Teeth large, very powerful, the cutting edges of the mandibulars more concave than those above, and the suture with a pronounced groove but no ridge on either side. Bifid tentacles large, very conspicuous, and placed about the last third of the space between the upper lip and the front margin of the eye. Interorbital space broad and slightly convex. Gill-opening equal to the internasal space or not quite as wide as the base of the pectoral. Peritoneum silvery. Lips, eyehds, margin of gill-opening, bases of the fins and the caudal peduncle naked, the other portions covered with rather large sharp spines, those on the belly longest and more or less slightly clavate. Origin of the dorsal fin is much nearer the posterior margin of the eye than the tip of the caudal fin, and the fifth ray is the highest. The anal is similar to the dorsal, entirely posterior, and with the sixth ray the longest. Caudal long, rather broad, and with the margin slightly rounded. Pectoral broad, the upper rays the longest, and the fin rather short. All the fins have rounded edges. Color in alcohol dark brown, the belly paler, and everywhere more or less uniform. Dorsal, anal and pectoral uniform pale brown. Caudal dark brown, with the margin all around broadly bordered with blackish brown. Each prick or spine is pale or whitish. The lips and bases of the fins are very dark brown. Length 3f inches. One example. SCORP^NID^. 5. Pterois lunulata Schlegel. Plate VI. 1843. Pterois lunulata Schlegel, Fauna Japonica, Poiss., p. 45, Plate 19; la bale de Nagasaki, principalement a rembouchure des fleuves. 1866. Pterois lunulata Playfair, Fishes of Zanzibar, p. 48; Aden; Zanzibar. [No descr.] Head 2y%; depth 3i; D. XIII, I, 10; A. Ill, 7; P. 14; V. I, 5; scales 64 in a lateral series to the base of the caudal, and about 4 more on the basal portion of the latter ; about 1 1 scales between the second dor- sal spine and the lateral line; 16 scales between the lateral line and the origin of the anal, and about 25? between the former and the middle of the belly; width of head 1^ in its length; depth of head 1^; snout 2| in head; eye 4^; maxillary 2; mandible If; interorbital space 4|; least depth of caudal peduncle 3^; first dorsal spine If; second 3 in body; hird 2f; fourth 2^; fifth 21; sixth 2i; thirteenth 2f in the head; 1903.] NATURAL SCIENCES OF PHILADELPHIA. 167 third dorsal ray, H; third anal spine 1|; third anal ray 1; caudal 2f in body; ventral spine 1§ in head; ventral fin 2^ in body. Body elongate, compressed, the greatest depth about the anterior dorsal spines. Head compressed, rather deep and somewhat bluntly pointed. Its greatest v^ndth is a little greater than the greatest width of the body, which is at the bases of the pectorals. Snout rather long, blunt, and with a marked prominence above. Preorbital space broad. Eye moderate, high, and a little anterior in the head. Mouth large, oblique, and the maxillary reaching back till below the front rim of the pupil. Distal expanded extremity of the maxillary broad, equal to three-quarters the eye-diameter, and inclined below ventrally. Jaws and vomer with bands of very fine teeth. Tongue thick posteri- orly, and in a rather long, thin, compressed point, which is free in front. Lips fleshy, the lower on the side of each ramus of the mandible very thick. The lower jaw large, protruding when the mouth is closed. and with a small protuberance at the symphysis. The interorbital space is rather narrow, deeply concave, and with a rather shallow median groove. Nostrils well separated, circular, the rims a little elevated, and the anterior with a small thin flap. No nasal spines. Preocular spines very blunt, and more or less con- cealed in the skin. Postocular spine very broad, obttise. Coronal and tympanic spines developed as low obscure ridges, and with a simi- lar median spine between the former. Parietal spines large and broad. Nuchal spines small. Margin of the preorbital with two blunt, obtuse spines. Cheek with a bony ridge, furnished with a blunt spine at the preopercle. Margin of the preopercle with three obtuse spines. Mas- toid and suprascapula each with a long spine. Opercle with a broad spine posteriorly. Gill-opening large, continued forward below the posterior nostril. Rakers short, thick. Gill-filaments much longer, though also rather short. Pseudobranchiffi. Isthmus broad, with a broad groove, and the branchiostegal membranes connected for a short distance. Scales small, cycloid, and present on the greater part of the head, except the jaws, snout and maxillary. A pair of short narrow cutane- ous flaps at the tip of the snout, and the two large obtuse spines along the preorbital margin each concealed in a rather long flap. Three short dermal flaps along the lower margin of the preoperculum. No flaps at bases of the paired fins. Scales rather large on the side and posterior portion of the trunk. Bases of the caudal, pectoral and ventral with small scales, the other fins without scales. Lateral line 168 PROCEEDINGS OF THE ACADEMY OF [Feb. concurrent with the back, slightly inclined, and running out on the basal scales of the caudal. The tubes elongate. Peritoneum pale or whitish. Spines of the dorsal very high, graduated to the fourth, after which, to the eleventh, they are more or less equal. The last two are much shorter than the first. The membranes are only present on the basal portion of the fin, though higher below the ]:)osterior longer spines. Soft dorsal nearly twice as high as its base. Anal inserted nearly midway between the origin of the ventral and the Imse of the caudal. The anal spines are graduated to the third, which is the longest, and the membrane joining it with the first anal ray is complete. Soft anal long, inserted about under the origin of the soft dorsal, and twice as high as the length of its base. Caudal \'ery long, rounded and the middle rays the longest. Pectoral very long, the upper rays reaching beyond the caudal, and all united by a membrane, which is rather broad basally though not extending beyond half the length of the longest rays. Ventral inserted a trifle in advance of the pectoral, the spine slender, and the fin reaching a little beyond the origin of the soft anal. Color in alcohol pale brown, the head and trunk marked with a series of deep brown alternating narrow and Ij'oad vertical stripes, or l)ands. On the front part of the head these radiate from the eye, and those on the trunk extend \\\) on the l)asal portions of the vertical fins, though soon fading away on the soft dorsal and anal. Stripes on the interorbital space narrow and longitudinal. Region above the l-)ase of the pectoral more or less blackish. Spines of the dorsal marked with broad blackish cross-bands. Soft dorsal, anal and caudal, with the rays marked with about six series of narrow dark brown or l^lackish cross-bars. Pectoral more or less ])lackisli, the ra3's with broad whitish cross-bands, and reflected' on the membranes at the base of the fin as intlistinct pale blotches. The basal portion of the fin is pale brown, like the general body color, marked with deep l)rown bands running into the black membranes. Inside of the base of the pectoral blackish, with se^'eral paler bands, and a few white spots. A'entral blackish, with several series of small white spots. Length 4y\ inches. One example. I follow Colonel Playfair in recording this species from Zanzibar. Schlegel's figure is ver}' crude and might easily be mistaken for some other fish. Bleekei-'s figure is scarcely an imjirovement. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 169 CEPHALAOANTHID^. 6. Cephalacanthus spinarella (Linnaeus). 1758. [Gasterosteus] Spinarella Linnaeus, Syst. Nat., Ed. X, p. 297; India. ^ 1866. Dactylopterus orientalis Playfair, Fishes of Zanzibar, p. 49; Aden; Zanzibar. [No descr.] 1893. Dactylopteru!^ orientalis Pfeffer, Jahrb. Hamburg. 'NViss. Anst., X. p. 12; Sansibar. [No descr.] Head 3f ; depth 5; D. I-I-6-S; A. 6; P. 31 : A'. 5; scales, in a lateral series to the base of the caudal 50, and about 28 in a transverse series between the spinous dorsal and the middle of the belly ; width of head lyV in its length; depth of head H; snout 2 J; eye 2f ; maxillary 2f ; width of mouth 2|-; mandible 2f ; interorbital space lyo ; least depth of caudal peduncle 4+; first dorsal spine 1^; second 3^; third If; first dorsal ray 1^; base of soft dorsal H; base of anal 2f ; ventral l^. Body elongate, depressed, the greatest depth about the base of the ventral. Caudal peduncle broad, depressed, about as wide as deep at its least depth. Head large, broad, and the greatest width a little less than the great- est width of the body, which is at the bases of the pectorals. The upper profile forms obtusely above the front of the eye. Snout broad, blunt, rounded and projecting beyond the jaws. Eye large, nearer the tip of the snout than the gill-opening in profile, and with the marginal bones large and strong. Mouth inferior, much broader than long, and the maxillary hardly reaching below the middle of the orbit. Lips rather thick. Teeth fine, minute, and in bands in the jaws, none on the vomer or pa'atines. Tongue broad, short, angular, and not free from the floor of the mouth. Upper jaw produced beyond the mandible. Nostrils close together, with slightly raised rims, and the space be- tween the anterior pair, which are closer together than the others, about two-fifths the interorbital space. Interorbital space very broad and deeply concave. Surface of the head almost entirely rugose, and the opercle and preopercle striate. Suborbital spine short, reaching below, the upper margin of the preopercle, and forming a deeply dented margin running up to the lower posterior rim of the orbit. Preopercle ending in a \'ery long dagger-like spine reaching opposite the fourth dorsal spine. Xuchal crests strongly ridged, with several denticula- tions, and reaching posteriorly below the tip of the preopercular spine. Gill-opening rather small, lateral, its length al)out equal to the eye. Peritoneum pale. Scales strongly keeled, bony, and arranged in longitudinal series. -Based on [Ptingititis] pvsillus Linnaeus, Mvs. Adolph. Fred., 1754, p. 74, Plate 32, fig. 5 (tliree figures). [Xo locality.] 170 PROCEEDINGS OF THE ACADEMY OF [Feb., On the anterior portion of the liocly the series are close together, broadening out behind so that some above run up toward the middle of the back, ami some of tlie lower run toward the lower median line. On the chest the scales are small and only slightly keeled. On each lower side of the caudal pedunc'e are a series of three enlarged com- pressed scales. The keels on the scales, in this region, are also large. Base of the caudal above and below with elongate compressed scales. Bases of the other fins without any scales. First dorsal spine situated at the angle of the nuchal shield, a little before the gill-opening. Second dorsal spine short, nearer the third than the first. Third dorsal spine a little nearer the posterior margin of the orbit than the origin of the soft dorsal, and with the remaining spines all graduated to the last wliich arc ver}- short. Origin of the soft dorsal a little nearer the base of the first dorsal spine than the base of the caudal, and the anterior rays the highest. Anal inserted a little nearer the l^ase of the caudal than the base of the ventrals, and the second ray about equal to the length of the caudal peduncle. Caudal small, lunate. Pectoral long, reaching beyond the base of the caudal. A entrals close together, elongate, and reaching the anus, which is a little distance in advance of the origin of the anal. Color in alcohol brown, a little darker above. Back with four broad obscure saddle-like bands. A dark brown band connecting the eyes and continued down below on each cheek. First dorsal spine with a blackish membrane. Other dorsal fins dusky with obscin-e dark brown cross-bars. Anal and ventral pale brown. Caudal barred with brown. Pectoral blackish with obscure gray and dusky blotches. Length 4 inches. One example. BLENNIID^.' 7. Aspidontus tractus sp. nov. Plate VII Head 3f ; depth 5; D. XI. 28; A. 26: P. 14; V. 3: width of head 2yV in its length; depth of head ly%; snout 3i; eye 4^; width of mouth 4|-; ■' Graviceps, gen. nov. The genus Aspidontus as here understood includes but two species, the one described above and .4. iwniatus Quoy and Gaimard. The species included by Jordan and Snyder in their review of the Blennioid fishes of Japan (Proc. U. S. Nat. Mus., XXV, 1902, pp. 453 to 458) appear to form a separate genus. They are all easilj' distinguished bj- the short blunt snovit, which is not produced or pig-like. Type Petroscirtes elegans Stcindachner. {Grai'iceps, heavy-head.) 1903.] NATURAL SCIENCES OF PHILADELPHIA. 171 interorbital space 4; depth of caudal peduncle 2h; width of base of pectoral 4; length of pectoral 2\; ventral 2^. Body moderately elongate, strongly compressed, and the greatest depth about the middle of the belly. The greatest width less than that of the head, and at the bases of the pectorals. Caudal peduncle not defined, as the posterior rays of the dorsal and anal are more or less connected with the caudal, deep and strongly compressed. Head large, attenuate, conic, and its greatest width at the opercles. The upper and lower profiles nearly evenly straight, so that the tip of the snout is midway in the height of the head. Snout long, conic, produced, and not Very broad. Eye rather small, a httle superior, and well anterior. :\Iouth inferior, some Httle distance from the tip of the snout, very broad, transverse, and the gape not reaching the front margin of the orbit. Lips thin, that on the mandible rather broad. Teeth compressed, uniserial in the jaws, and attenuately rounded at their extremities. A large curved canine on each side of the mandible, and not separated from the others. No teeth on the vomer or palatines. Nostrils well separated, the anterior about the last fourth of the space between the tip of the snout and the posterior, which is over the front rim of the orbit. The internasal space is about two-thirds the diameter of the eye. Interorbital space rather broad, convex like the rest of the upper surface of the head. The suborbital bones are somewhat rugose and covered with a thin membrane. The sides of the head posterior and superior to the eyes are similar. The lower surface of the head is broad, convex and rounded like the top or upper surface. Gill-opening small, in front of the upper base of the pectoral and about equal to the eye. Peritoneum whitish. Anus near the origin of the anal fin. Body naked, the skin smooth. Head with a number of small mucous pores. Paired fins without flaps or sheaths. Origin of the dorsal well forward, only just a trifle behind the eye. The spines are very flexible, hardly to be distinguished from the rays, which are not branched. The bases of the former are a little further apart than those of the rays. The first few spines are graduated, after which the whole dorsal fin assumes a more or less uniform height to the last rays, which decrease and are united with the caudal peduncle by a low membrane reaching to the base of the caudal. The anal is similar to the dorsals, except that the margin of the fin is incised between the ravs. The anal spine is scarcely to be distinguished from 172 PROCEEDINGS OF THE ACADEMY OF [Feb., the rays. The origin of the anal is much nearer the base of the caudal than the tip of the snout. Caudal broad, somewhat truncate, and the lower rays the longest. Pectoral broad, also rather truncate. Ven- trals close together, inserted about opposite the last third of the space between the posterior margin of the eye and the margin of the gill- opening. Color in alcohol more or less dark slaty-blue, giving place to dark brown. The posterior half of the trunk shading into dusk3% A black band, at first about half or a little less than the e\^e-diameter, running from the snout through the eye and back along the upper part of the side out on the caudal till near the tips of the lower median rays, and then curving down or bent back a trifle. From above the origin of the anal this band becomes much broader, occupying most of the upper portion till about as broad as the interorbital space. Belly and lower surface rather pale. A short dark brown oblique streak before the base of the pectoral. Margins of the dorsal, anal and caudal whitish, that of the latter very broad, especially at the corners. On the former two fins there is a narrow submarginal ' blackish band, fading below toward the basal portions of these fins into the bluish body color. Pectoral and ventral pale Ijrown. Length 3^ inches. Type No. 24,207, Acad. Nat. Sci. Phila. Zanzibar. Coll. Sir Charles Eliot. The only example I have seen is described above. It is closely related to Aspidontus taeniatus Quoy and Gaimard. It differs, however, in the short dark streak in front of the base of the pectoral. Their figure is very rough, and shows the outer portions of the upper and lower caudal rays bluish.* Dr. Giinther's figiu'e of Petroscirtes tceniatus.'' if at all accurate, is c ertainly distinct. {Tractus, a streak.) ANTENNARIIDJE. 8. Antennarius argus sp- nov. Plate Vni. Head (measured \o axil of pectoral) 2f; depth 1; D. I-I-I-12; A. 7; P. 10; V. 5; width of head (measured to axil of pectoral) l^ in its length; snout 4^ in head; eye about S; maxillary 2; mandible 2; width of mouth If; interorbital space 3; bait 2\; second spine 3^; ■* Voyage de V Astrolabe, Zool., 1S34, p. 719, Plate 19, fig. 4; I'ile de Guam, devaiit la ville d'Agagna. •=• (Fische der Siidsee) Journ. Mus. Godef., XI, 1876, p. 195. Plate 114, fig. A; Freundschafts-, Paumotu- und Fidschi-Inseln. 1903.] NATURAL SCIEN'CES OF PHILADELPHIA. 173 third spine If ; fourth dorsal ray 1| ; base of anal 2^ ; fourth anal ray 2| ; caudal 1^; least depth of caudal peduncle 3^; base of pectoral 2+; base of ventral 3^. Body deep, compressed, and the greatest depth about midway be- tween the base of the first dorsal spine and the first dorsal ray. The back is elevated and the belly round and swollen. Caudal peduncle compressed, rather small. Head very large, deep, with a steep declivous profile. The lower sides of the head about the greatest depth of the body, also its greatest width. Snout very short, steep, rough and broad. Eye small, very high and anterior. Mouth very large, nearly vertical, and the lower posterior margin of the maxillary hardly reaching below the front margin of the eye. Mandible very broad and with a slight knob at the symphysis. Lips rather thick. Teeth minute, and in rather broad bands in the jaws. Vomer and palatines with bands of minute teeth. Tongue large, thick, occupying the whole of the floor of the mouth, and with a patch of large^'short coarse teeth medianly. Nos- trils small, obscure, close together, and near the edge of the snout. Interorbital space broad, and rough. Top of the head below the third dorsal spine swollen. Gill-opening a small pore at the lower base of the pectoral, and nearty in the middle of the length of the entire fish. Peritoneum silveiy. Anus near the front of the' anal fin and with a small genital papilla posterior. Body very finely roughened everywhere, except on the lower surfaces of the pectorals and ventrals. Tubercles on the head, and in the lateral line anteriorly, not very conspicuous. Bait with a large tuft' of filam.ents at its extremity. Second dorsal spine closely following, and united behind with a broad roughened membrane with the interorbital space. Third dorsal spine clavate beginning at the tip of the|;depressed second spine or a little behind the eye, and also united to the occiput with a broad rough membrane. Soft dorsal rather high and long, and its margin straight. Anal small, very posterior, and extending back rather far on the caudal peduncle. Caudal large, broad and the upper rays the longest. All of the vertical fins rounded. Pectoral large, and the tips of the rays projecting a little beyond the membranes. Wntral similar to pectoral. Color in alcohol pale or dusky gray-brown, more or less uniform, and marked everywhere with small round black ocelli, many scarcely larger than the pupil of the eye. The lower surfaces of the pectoral 174 PROCEEDINGS OF THE ACADEMY OF [Feb., and ventral are blackish without distinct spots. On the back and head many of the ocelH become very small. The outer portions of the paired fins are also darker than the other parts. Along the upper portion of the soft dorsal are four large blackish ocelli arranged at equal distances, and posteriorly on the base of the fin a much larger one. Anal similar but the basal ocellus at the fifth and sixth rays. There are also several enlarged ocelli on the sides of the body and caudal fin. Length 3/g inches. Type No. 24,208, Acad. Nat. Sci. Phila. Zanzibar. Coll. Sir Charles EHot. One example, described above. It resembles Antennarius phyma- todes Bleeker somewhat in color, but is apparently more closely re- lated to Antennarius polyophthalmus Bleeker, from which it is distin- guished by the very numerous ocelli. Antennarius nigromaculatus Play fair is said to have a large black patch extending over the whole of the abdominal region, {Argus, hundred-eyed, on account of the very numerous ocelli.) 9. Antennarius tuberosus Cuvier. 1817. Chironectes tuberosus Cuvier, Mem. Mus. Hist. Nat. Paris, III, p. 432; risle de France. (Coll. M. Mathieu.) Head (measured to axil of pectoral) 2^; depth 2; D. I-I-I-ll; A. 7 ; P. 9 ; V. 5 ; width of head (measured at opercle) about li in its length; snout 4 in head; eye about 6; maxillary If; width of mouth If; interorbital space 3^; bait 4f ; second spine 4.V; third spine 2f ; eighth dorsal ray 2; base of anal 1^; fourth anal ray 2; length of caudal 1^; least depth of caudal peduncle 2^; base of pectoral 2f ; base of ventral Body elongate, compressed, and the greatest depth about the origin of the soft dorsal. The back is elevated, especially between the third dorsal spine and the origin of the soft dorsal, sloping down gradually behind to the caudal peduncle. The abdomen is very large, rounded, or swollen. Caudal peduncle small, and but little free. Head moderate, very deep, and with a steep round anterior profile. Side of the head more or less swollen in appearance, but with the greatest width of the body at the base of the pectoral in front. Snout very short, blunt, broad, and a little inclined. Eye small, very high and anterior. Mouth large, inclined forward so that the chin is pro- duced. Maxillary concealed below. Mandible very broad and with a small symphyseal knob. Lips thick. Teeth small and in rather broad bands in the jaws. Vomer and palatines with bands of fine teeth. Tongue large, free around the edges so that it is very thick, 1903.] NATURAL SCIENCES OF PHILADELPHIA. 175 occupying the whole of the floor of the mouth, and with a patch of rather large teeth medianly. Nostrils small, close together and a little nearer the edge of the mouth than the front margin of the eye. Interorbital space moderate, elevated and rough. Top of the head below the third dorsal spine not very much swollen. Gill-opening a small pore just below the base of the pectoral, and nearer the edge of the chin in front than the base of the caudal. Peritoneum grayish-brown . Anus conspicuous, in front of the origin of the anal. Body very finely roughened everywhere, except on the extremities and lower surface of the pectoral and ventral rays. Tubercles on the head and anterior portion of the lateral line obscure and inconspicuous. Bait short, with a small tuft of filaments at its extremity . Second dorsal spine closely following, about opposite the front rim of the orbit, free, and without membrane behind. Third dorsal spine large, thick, not very high, and joined to the occiput by a narrow membrane. Soft dorsal long, beginning a little behind the base of the pectoral above, and with the upper margin only slightly incised between the rays. Anal posterior, beginning a little nearer the tip of the caudal than the origin of the ventral, rounded, and the edge more or less similar to that of the soft dorsal. Caudal rounded, somewhat ex- panded, and the median rays the longest. Pectoral low, and the rays free for a good portion of their extremities. Ventrals situated below the third dorsal spine and the ends of the rays also free for a good dis- tance. Color in alcohol dark gray-brown, the margins of all the fins rather broadly whitish. Side and head more or less obscurely marbled with darker brown. The outer submarginal portions of all the fins blackish, variegated with paler or brownish. Base of the caudal pale brown. Length 1| inches. One example. The name bigibbus'' cannot ilate from the original edition of Lacepede, as it was not used as a binomial. The next name apparently available is that of Cuvier. Fauxal Works, 1865. R. L. Playfair and A. C. L. G. GrxTHER. The Fishes of Zanzibar. Lon- don. Pp. 153, Plates 21. (Acantliopterygii by Lieut-Col. Playfair and Pharyngognathi, etc., by Dr. Giinther.) ^ Lophie double-bosse Lacepede, Hist. Nat. Poiss., I, 1798, pp. 302, 325. [No locality. From Commerson MSS. as " Aiitcnnarius bigibbus, nigro et griseo variegatus."] 176 PROCEEDINGS OP THE ACADEMY OF [Feb., 1S68. A. C. L. G. GrxTHER. Additions to the Ichthyological Fauna of Zanzi- bar. Ann. Mag. Nat. Hist. London, (4) I, pp. 457-459. with fig. W. Peters. Cliiloglanis Deckenii, und einige andere Susswasser-fische aus Ostafrika. Monats. Ak. Wiss. Berlin, pp. 598-602, ^yith plate. 1869. R. L. Playfair. Further Contributions to the Ichthyology of Zanzi- bar. Proc. Zool. Soc. London, pp. 239, 240. Addendum", p. 241, by Dr. Giinther. 1873. A. C. L. G. GiiNTHER. Further Additions to the Ichthyological Faima of Zanzibar. Ann. Mag. Nat. Hist. London, (4) XII, p. 182. 1889... Georg Pfeffer. Ubersicht der von Herrn Dr. Franz Stuhhnann in Agypten, auf Sansibar und dem gegeniiber liegenden Festlande gesammelten Reptilien, Amphibien, Fische, Mollusken und Krebse. .Jah-b. Hnmbarg. iciss. Anst., VI, Fische, pp. 13-23. 1893. . Ostafrikanische Fische gesamnielt von Herrn Dr. F. Stuhlmann im Jahre 1888 und 1889. Jahrb. Hamburg. iviss. Anst., X, pp. 1-49, Plates I-III. 1899. R. Jatzow and H. Lenz. Fische von Ost-Afrika, Madagaskar und Aldabra. Abhand. Senck. Naturj. Gesel, XXI. pp. 497-531, Plates .34-.3lack blotches on each side. Distribution.- — This species is apparently rather generally distributed over Texas. Specimens in the material examined are from Shovel Mount, Burnet county; Eound ]>Iountain, Blanco county; Austin, Travis county; San Marcos, Hays county; Devils River and Pecos High Bridge, ^^alverde county. The U. S. National Museum collection also contains specimens from Victoria, Mctoria county; Dallas, Dallas county, and Carrizo Springs, Dimmit county. The only definite record from New Mexico is from Las Cruces, Donna Ana county. The species 1903.] NATURAL SCIENCES OF PHILADELPHIA. 187 is seen to extend from east central Texas to the Rio Grande ^' alle}- in New Mexico, and from the Red River region to south central Texas. Remarks. — ^This species exhibits a great A-ariation in size, and also in the shape of the pronotum. Specimens Examined. — Nineteen males, eight larvae: 14. Shovel Moimt, Burnet county, Tex. September 2-October 29, 1901. (Sehaupp.) (Acad. Nat. Sciences Phila.) 4. Round Mountain, Blanco countj^. Tex. (Sehaupp.) (Acad. Nat. Sciences Phila.) 2. Austin, Travis county, Tex. Feb. 19, 1903. (Wheeler.) (Acad. Nat. Sciences Phila.) 1. San Marcos, Hays county, Tex. (Pilsbry.) (Acad. Nat. Sciences Phila.) 2. Devils River, Valverde county, Tex. (Pilsbr}'.) (Acad. Nat. Sciences Phila.) 2. Pecos High Bridge, Valverde county, Tex. (Pilsbry.) (Acad. Nat. Sciences Phila.) 2. Texas. Riley Coll. (U. S. N. M.). Homoeogamia erratica n. sp. Type. — ; Prescott, Yavapai county, Ariz. June 10, 1902. (Oslar.) (Acad. Nat. Sciences Phila.) Differing from bolHana in the smaller size, the greater interspace between the eyes (which is much more than half that between the ocelli, instead of half or less than half), the obscure character of the nervures of the distal portion of the marginal field, and more forked structure of the nervures of the distal portion of the discoidal field. Description. — d^. Size rather small; form acute elliptical; surface sub-coriaceous. Head with the interspace between the eyes slightly less than that between the ocelli ; ocelli very large, and in contact with the eyes along their superior margins ; antennae almost equal to the bod}' in length ; basal joint rather elongate. Pronotum sub-triangular, the basal line being represented by the posterior margin; posterior margin very broadly arcuate, lateral angles rounded; anterior margin and surface covered with hairs, those on the margin being rather long, and those on the surface very short. Tegmina elongate, the apex well rounded; proximal portion of the anterior (lateral) margins sparsely haired ; basal field ■\nth the nervures but shghtly marked distally, not visible in the proximal portion; nervures of the discoidal field parallel^ oblique, straight, frequently bifurcate ; anal sulcus rather deep, strongly arcuate. Wing reaching the tips of the closed tegmina, apex broadly rounded. Supra-anal plate produced, -with a median triangular emar- ISS PROCEEDINGS OF THE ACADEMY OF [Feb., gination. Sub-genital plate produced, rounded, with a shallow, rotundate, apical emargination. Cerci short, depressed, not exceed- ing the sub-genital plate in length. General color brownish-gray, the pronotum with a few median marks of dull brown. Tegmina obscurely dusted with irregular blotches of brownish. Head black, except the labrum, clypeus, mandibles and inter-antennal area which are yellowish-white ; ocelli very pale amber ; antennae pale brownish, except the basal joint which is yellowish- white. Limbs and under surface very pale sandy brown; spines on the former rich reddish-brown. Length of iDody 14.5 mm. ; length of tegmina 18 mm. ; greatest width of tegmina 7 mm.; length of pronotum 4.5 mm.; width of pronotum 6.2 mm. 9 . Unknown. Distribution. — This form is known only from Prescott and Tucson, Ariz. Remarks. — This species presents some considerable variation in size, and also in the intensity of the coloration, The specimens from Prescott are much grayer than the Tucson specimen, which is a pale straw-brown in color. Specimens Examined. — Three males : 2. Prescott, Yavapai county, Ariz. June 10, 1902. (Oslar.) (Acad. Nat. Sciences Phila.) 1. Tucson, Pima county, Ariz. June 22. (U. S. N. M.) Homoeogamia apacha Saussure. 1S93. [Homceogamia] apacha Saussure, Revue Suisse de Zoologie, I, fasc. 2, p. 296. [Chihuahua, Mexico.] 1894. Homoeogamia apacha Saussure and Zehntner, Biol. Cent.-Amer., Oi-thopt., I, p. 107. [State of Chihuahua, Mexico.] Description .—(J . Size small; form elongate ovate: surface sub- coriaceous. Head wdth the interspace between the eyes equal to that betw^een the ocelli; antennse not quite equal to the body in length. Pronotum triangular, the posterior margin longer than the other sides ; surface covered with short appressed hair, the anterior margin with them much the longer. Tegmina elongate with the apex well rounded; basal field comparatively broad, the nervures distinctly visible in all but the extreme basal portion; nervures of the discoidal area rather irregular, frequently forked. Wings with the apex sub-acuminate. Supra-anal plate produced, rounded, deeply divided apically. Sub- genital plate produced, apically with a broad, shallow, triangular emar- gination. Cerci shghtly longer than the sub-genital plate, appressed acuminate. Anterior metatarsi longer than the tibia; median meta- 1903.] NATURAL SCIENCES OF PHILADELPHIA. 189 tarsi not more than two-thirds the length of the tibia; posterior meta- tarsi sUghtly more than half the length of the tibia. General color wood-brown, marked on the pronotum with a darker shade of same color; hair on the pronotum red-brown. Tegmina obscurely mottled with general tint, the lighter spots being sub-hyaline, except in the anal and basal section of the discoidal field. Color of the imder surface and limbs pale wood-brown, darkest in the anal region and on the spines of the limbs. Length of body 11.7-15.2 mm.; length of tegmina 14-18; greatest width of tegmina 6-7; length of pronotum 3.7-4.5; width of pronotum 5.2-6.5. 9. "(Nymph?) Rufo-ferrugineous, apterous, head and feet testa- ceous. Supra-anal plate transverse, margin Uttle arcuate, slightly incised. Cerci tuberculiform. Last ventral segment flat, rounded, both sides sinuate, produced in the centre. Length 13; pronotum 4.6; width of pronotum 7 millim." Distribution. — The series examined includes specimens from Phoenix, Tempe and Fort Grant, Ariz. These with the type locality, Chihuahua, cover the known distribution of the species. Remarks. — The most striking variation in the specimens examined appears to be that of size. The coloration appears, from the speci- mens at hand, to be quite constant. Specimens Examined. — Five males : 2. Fort Grant, Graham county, Ariz. July 17 and 22. (Hubbard.) (U. S. N. M.) 2. Phoenix, Maricopa county, Ariz. April 29, 1902. (Oslar.) (Acad. Nat. Sciences Phila.) 1. Tempe, Maricopa county, Ariz. April 26, 1902. (Oslar.) (Acad. Nat. Sciences Phila.) Subgenus EKEMOBLATTA n. subg. Subgeneric Characters. — c?. Size small. Vertex of the head with a distinct angular transverse ridge. Tegmina with the nervures in the anal field very distinct. Median and posterior tibiae with seven apical spines. Supra-anal plate very broadly emarginate apically. 9 . Unknown. Type. — Homceogamia subdiaphana Scudder. Homoeogamia subdiaphana Scudder. 1902. Homoeogamia subdiaphana Scudder, Proc. Davenport Acad. Sci., IX, p. 19. [Las Cruces, N. M.] 1902. Hornoeogamia subdiaphana Rehn, Proc. Acad. Nat. Sciences, Phila., LIV (1902), p. 717. [Alamogordo and HighroUs, N. M.] c?. Size small; form slender and elongate. Head with the inter- 190 PROCEEDINGS OF THE ACADEMY OF [Feb., space between the eyes slightly greater than that between the oceUi; vertex with a distinct transverse ridge which divides the apical from the inferior surface of the head ; antennae slender, filiform, shorter than the body. Pronotum sub-triangular; surface velvetinous and some- times haired; the posterior margin broadly arcuate, the anterior semicircular; both margins closely haired, those of the anterior margin much longer than on the posterior; lateral angles obtuse, inconspicu- ous. Tegmina with the surface glabrous, the distal half membrana- ceous; apex sub-acuminate, rounded; anterior (lateral) margin basally supplied with long hairs ; marginal field with the nervures absent in the proximal portion; discoidal field with the nervures obliquely parallel, almost longitudinal; anal field with the nervures distinctly visible and strongly arcuate. Tarsi with seven apical spurs. Sub-genital plate with the apex broadly emarginate, the lateral shoulders pro- duced into distinct dentiform processes, the intervening diastema being of moderate depth and evenly rounded. Cerci flattened, very short. General color wood-brown. Head blackish; ocelli aml^er-yellow. Pronotum with the anterior and lateral margins pale yellowish, the hair, which is present on the margins and sometimes on the disk, being of the same tint; disk blackish with two large lateral and one small medio-posterior dot of red, which are more or less conspicuous. Teg- mina with the general tint strongest on the nervures of the proximal half of the tegmina. Under surface yellowish-brown, suffused with darker brown in the anal region. Length of body 9.5 mm.; length of tegmina 14 mm.; greatest width of tegmipa 5 mm. ; length of pronotum .3-3.5 mm. ; width of pronotum 4.2-4.5 mm. 9 . Unknown.^ Larva: Form ovate, the abdomen very broad. Pronotum similar to that of the male in general outline. Supra-anal plate transverse, the posterior margin rounded and centrally emarginate. General color wood-brown, edged on the pro- and mesonotum wdth ochraceous. Pro-, meso- and metanotum centrally ornamented with blotches of ochraceous, the penultimate abdominal segment bearing lateral blotches of the same tint. Distribution. — This species has been recorded only from the Rio Grande Valley at Las Cruces, N. M., and the San Augustine Plain * The specimen described by the author (Proc. Acad. Nat. Set. Phila., LIV (1902), p. 717) as a female is in a larval condition and the sex is uncertain. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 191 and foothills of the Sacramento [Mountains at Alamogordo and High- rolls, Otero county, N. M. Remarks. — This species is so peculiar and distinct from the other forms of the genus that it is easily recognized. The series examined is quite uniform in size, and the color differences are very slight. Speci?nens Examined. — Eleven; ten males, one nymph: 10. Alamogordo, Otero county, N. M. April 11-June 6, 1902. (Vie- reck and Rehn.) (Acad. Nat. Sciences Phila.) 1. Highrolls, Otero county, N. M. June 13, 1902. (Viereck.) (Acad. Nat. Sciences Phila.) Unidentified Forms. Polyphaga (Homeogamia) brasiliana Saussure. 1864. Pobjplmga (Homeogamia) Brasiliana Saussure, Mem. I'Hist. Nat Mexique, IV, p. 228. 1868. Pohjphaga Brasiliana Walker, Catal. Blatt. Brit. Mus., p. 15. "Nymphe d"'. Corps un peu voute, suborbiculairc. finement striolc ou ponctue. Le premier arcticle des antennes mediocrement long, assez gros; le deuxieme petit; le troisieme deux fois plus long. Pro- thorax large et court; son bord anterieur a courbure obtuse, subangu- laire an milieu; les angles lateraux tres-aigus; le bord posterieur un peu arque, subsinne avant les angles. Abdomen aplati, large, a bords faiblement dentes; segm.ents dorsaux 8, 9 tres-distincts, decouverts; le bord posterieur du 7me presque en demi-cerclc; en dessous le penul- tieme segment petit, tres-arque; ses deux extremites presque cachees sous le precedent. Plaque sous-genitale ayant son bord posterieur en forme d'angle obtus; ce bord fortement depasse par la lame sous- anale fentlue, laquelle est a son tour debordee par la plaque suranale ; celle-ci transversale, a bord posterieur presque droit, et un peu fendu; la face superieure cannelee au milieu. Styles anaux tres-distincts. Pattes greles, allongees; epines tibiales longues et greles. "Couleur d'un marron ferrugineux, presque testacee en dessous; vertex, front et labre superieur, offrant une bande brune; antennes ferrugineuses, plus foncees en dessus; bords du thorax cilies de polls ferrugineux, et varies de testace-jaunatre; sur la ligne mediane chaque segment orne de deux joints jaunes; ces points se continuant sur I'abdomen en devenant plus vagues; les segments abdominaux ornes en outre de chaque cote d'une marque jaune et d'un point brun; te thorax en dessous jaune-testace. ''Longueur du corps, 0,017; — largeur du prothorax, 0,0115; — id. de I'ahdomen, 0,0145. ''Hahite: Le Bresil." 192 PROCEEDINGS OF THE ACADEMY OF [Feb., Polyphaga ae^ualis Walker. 1871. Polyphaga aqualis Walker, Catal. Spec. Derm. Salt. Brit. Mus., V, Suppl. Catal. Blatt., p. 3. "Male. Piceous. Head black; clypeus, palpi and antennse red. Prothorax scabrous, transversely elliptical; anterior half of the border red; some longitudinal impressed lines in the disk. Pectus, legs, sides of the abdomen and hind borders of the segments tawny; femora setose beneath. Fore wings extending for full half the length beyond the abdomen, with numerous minute pale testaceous dots, occasionally with a large pale testaceous patch in the disk near the base. Hind wings pellucid ; veins paler. Length of the body 8-9 lines ; expansion of the fore wings 28-30 lines. The prothorax is regularly elliptical, and thus differs from that of P. Mexicana, in which the hind part is broader than the fore part, and is truncated on each side. a-b. South Mexico." This is apparently a member of the genus Homceogamia, but cannot be positively identified. It is possibly a synonym of mexicana, as Walker's knowledge of the latter species must have been limited, as his statement regarding the form of the pronotum in that species is partially erroneous. Homoeogamia n. sp. Bolivar. 1890. Homceogamia n. sp. Bolivar, Ann. Soc. Ent. France (6), X, p. 137. "Un seul exemplaire a I'etat de larve. "Possedant dans ma collection des individus a divers-degres de developement, je pens affirmer seulement que cette espece ne se rapport pas a H. mexicana Burm., avec qui je I'ai comparee, ni a H. hrasiliana Sauss., bien que cette derniere espece ait ete decrite sur un male a I'etat de nymphe. Le pronotum est convert de granulations mili- aires, ses sillons dorsaux etant remplaces par de simple depressions, ce qui donne a I'insecte un facies un pen different des autres Homceo- gamia. — Caracas. " Je crois cette espece nouvelle, mais ne possedant pas des individus bien developpes, je ne la decris pas." 1903.1 NATURAL SCIENCES OF PHILADELPHIA. 193 MOLLUSCA OF WESTERN ARKANSAS AND ADJACENT STATES, WITH A REVISION OF PARAVITREA. BY HENRY A. PILSBRY. Early in 1901 Mr. J. H. Ferriss explored for mollusks a portion of southwestern Missouri, western Arkansas, and some adjacent locali- ties in the Indian Territory, the region covered lying mainly north of that exploited in 1900.^ The rugged topography of a portion of this region, and its elevation above the plains on all sides, have resulted in the evolution of many species and subspecies special to the tract. The general conditions of life are varied from those prevailing over the comparatively level States along the Mississippi, and with this change has come readjust- ment on the part of the snails. The mode of this readjustment I hope to study further when more material is available. From the data pre- sented in this paper and my article of 1900 on the same fauna, it appears that — (1) Species having a wide geographic range become much more vari- able in this district. (2) Variation is usually not indiscriminate, hut in the large majority of individuals follows one or two definite lines of deviation from the "nor- mal" or widespread form. (3) Various specific stocks show various degrees of deviation or differentiation from their widespread "normal" forms. The first of these propositions is exemplified by a large nmnber of species, and would be noticed by any one possessing the material for comparison. The second point is one of some importance. It is illustrated likewise by many species, of which several may be taken as examples. Gastrodonta demissa is a snail ranging from "Western Pennsylvania to Georgia, west to Arkansas and Eastern Texas. "- 1 See Pilsbry in these Proceedings for 1900, pp. 449-459, and especially Ferriss. Nautilus, XIV, pp. 25-31, July, 1900, where the country is described, and manj^ species not in other lists are noted. Information bearing on the molluscan fauna of adjacent districts may be found in the following papers: F. A. Sampson, "Shells of Pettis county, Mo.," in Bull. I, Sedalia Natural History Society (1885) : "Preliminary List of the Mollusca of Arkansas," Ann. Rep. Geol. Survey of Arkansas for 1891, Vol. II, pp. 179-199 (1893); C. T. Simpson, "Notes on Some Indian Territory Land and Fresh-water Shells," in Proc. U. S. National Mu- seum, 1888, pp. 449-454. ■^ Pilsbry, Catal. of the Land Shells of America, p. 28 (1898). 13 194 PROCEEDIXGS OF THE ACADEMY OF [Feb., In this area it is not known to vary much east of the Mississippi. Several years ago I received a number of specimens from Arkansas Avhich differed from all previously known in having the whorls so closely coiled as to make the axis quite imperforate, and at the same time the whole shell was flatter. This form I called Gastrodonta hrittsi. Subsequently, specimens from another Arkansan locality came to my hands, in which there was a lamella within the aperture, a structure unknown in the species elsewhere. These became G. demissa var. lamellata. The abundant series collected by Mr. Ferriss since these supposed subspecies were discriminated, shows that in their area they coexist with snails indistinguishable from Eastern G. demissa, and that the variation of that species in this area is not indeterminate, but toward the one or the other of these two diverse modes of modification. Polygyra monodon is in the same case. In western Arkansas, three weakly differentiated forms occur, which have been named alici(B, friersoni and imperforata. All are closely related to, and doubtless descendants of, the widespread P. m. jraterna; and all three occur around some localities.^ But in examining several hundred indi- viduals I did not find the variation indiscriminate, even in a small percentage of the snails, but always toward one or the other of the three differentiation-lines signahzed by the three subspecific names, without individuals intermediate between them, or otherwise modified, as tlie species is in some other regions. Without further multiplying instances, I may say that while in certain districts, usually of rugged and varied topography, widespread species of snails become locally variable, there has come to my notice but little of the multifarious variation called for by the theory of natural selection. What strikes one is the uniformity with which modifications, varying in amount, yet follow definite paths. This impression left by studies of the snails of the United States, I would never have gained from what I have seen of certain Antillean genera, such as Cerion or the Jamaican Urocoptis. In these groups there seems to be great and multifarious variation, without elimina- tion of intermediate forms, thus making specific lines merely arbitrary. List of Species. Localities marked with an asterisk (*) in this list are given on Mr. Ferriss' authority. ^ See my former paper on Arkansan snails, Proc. A. N. S. Phila., 1900, pp. -154-456, where the data on distribution are given. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 195 POMATIOPSIN^. Pomatiopsis lapidaiia (Say). Seligman, Barry county, Mo. HELICINID^. Helicina orbiculata tropica (Jan.). Seligman,* Mo. ; Petit Jean,* 'Yell county, Ark. HELICID^. Polygyra leporina (Gld.). Poteau* and Antlers,* I. T. Polygyra dorfeuilliana (Lea). Springfield, Green county, southwestern Missouri. Typical P. dorfeuilliana sampsoni Weth., 7.8 to 8.2 mm. diam. Seligman, Barry county, southwestern Missouri. Typical P. d. sampsoni, diam. 8-9 mm. Chester, Crawford county, Ark. Varying from 6| to 8 mm. diam., some specimens shomng transition toward var. sampsoni in width of the umbiUcus. Rich Mountain, Polk county. Ark. Specimens typical, 7-7f mm. diam. Hot Springs, Garland county. Ark. Like the preceding. Carrion Crow Mountain, in Gum-dog township, Pope county. Ark. This is about ten miles north of Petit Jean Mountains. Rather heavy specimens, 6.2 to 7.6 mm. diam., with the lower lip-tooth somewhat prominent in a basal view. Tushkahoma, I. T. The specimens vary from 6.2 to 8 mm. diam., are rather solid and somewhat more striate beneath than in typical dorfeuilliana. The parietal tooth is rather small. The umbilicus varies from that of typical dorfeuilliana to as wide as sampsoni, so that there is here a perfect transition between the two. A large series was collected. Antlers, Choctaw Nation, I. T. Three specimens measuring 6, 6.2 and 7.5 mm. diam. The lip is much thickened, immersing the teeth, and the parietal tooth is unusually large and square. In the largest shell the umbilicus is that of sampsoni, and the aperture less obstructed, diam. 8 mm. Standley, Choctaw Nation, I. T. Diam. 7 to 7.5 mm., the specimens varying from dorfeuilliana to sampsoni in umbilicus. Aperture normal. Poteau, Choctaw Nation, I. T. Rather solid, distinctly striate beneath, varving from 7.5 to 8.7 mm. diam.. and from the small 196 PROCEEDINGS OF THE ACADEMY OF [Feb., umbilicus of dorfeuilliana to the widely open condition of sanipsoni, most of the specimens intermediate in size and umbilicus. A large series. In some portions of southern Missouri and adjacent counties of Arkansas, the specimens are larger than typical dorfeuilliana, and widely open beneath, showing over one whorl. They are usualh* qviite glossy, and rather weakly striate on the base. This form, var. sampsoni Weth., seems to be quite distinct in this area. In the eastern part of Indian Territory, the shells are rather hea^aer, more distinctly striate on the base, and wherever a large series was collected they vary from the dorfeuilliana form, in which but one whorl or a little less is exposed beneath, to the sampsoni form of base; most specimens being intermediate in these characters. This area therefore is one in which differentiation has not taken place — a transition region. Cf. these Proceedings for 1900, p. 449. Polygyra jacksoni (Bland). Springfield, Green county, and Chester,* Crawford county. Mo.; Poteau* and Rich Mountain,* Ark. Polygyra cragini (Call). DeKalb,* Cleburne county. Ark. ; Antlers,* I. T. Polygyra neglecta Pils. T. fallax Say var. minor, Wetherby, Some Notes on American Land Shells. No. II, p. 11, in Joum. Cincinnati Soc. N. H., IV, December, 1881, p. 383. Polygyra neglecta Pils., Nautilus, XIII, p. 40 (August, 1899). The above reference to Mr. Wetherby's work was overlooked by me, or not recognized as pertaining to this species, when I defined P. neglecta in 1899. He says: "Years ago I received from Springfield. Mo., a small variety of this species [Triodopsis fallax], much lighter colored, with a thicker and heavier shell than the type, with the peristome reflected backward and rounded, and having a very distinct facies. I have recently received the same variety from Mr. Sampson, who collected it at Eureka Springs." The form would hardly be recognized from this note without speci- mens from the localities mentioned ; and as several forms of the genus have already been called "var. minor," it will probably not be advis- able to revert to that name. Mr. Ferriss procured specimens at Seligman, Barry county, and Springfield, Green county, both in south- western Missouri. Polygyra infleota (Say). Petit Jean, Yell county, Ark.; diam. 9-11 mm. Seligman, Mo.. normal specimens with strong teeth, diam. 10 mm., and a very distinct 1903.] NATURAL SCIENCES OF PHILADELPHIA. 197 race in which the teeth are much reduced, though still the Hp-teeth are decidedly stronger than in P. edentata. Specimens measure — Alt. 5.3, diam. 11.5 mm.; whorls 4f. Alt. 5, diam. 9.5 mm.; whorls 4^. The last whorl is somewhat more costulate behind the lip than usual in P. inflecta. This form may be called var. media. P. inflecta was taken also at Rich Mountain,* Hot Springs* and Carrion Crow Mountain,* Ark.; Tushkahoma* and Standley,* I. T. Polygyra edentata (Sampson). Chester, Crawford county, western Arkansas. Two specimens sent measure 12 and 14 mm. diam., and have 5 and 5^ whorls. Eighteen were taken at this locality. Polygyra albolabris alleni iWetherby). Mesodon albolabris Say, and var. minor A. G. Wetherby, Some Notes on American Land Shells, No. II, in Joum. Cincinnati Soc. Nat. Hist., IV, p. 11, December, 1881. Eureka Springs, Carroll county. Ark. M. albolabris Say, F. A. Sampson, Bull. No. 1, Sedalia Nat. Hist. Soc, p. 19, 1885. Sedalia, Mo. Mesodon albolabris Say. var. alleni (\Vetherby) and var. minor Sampson, MoUusca of Arkansas, in Ann. Rep. Geol. Surv. of Ark., II, pp. 189, 190, 1893. Carroll, Benton, Sebastian, Crawford, Garland, Washington, White, Johnson, Nevada, and Independence counties. Ark. Polygyra albolabris alleni Weth., Pilsbry, in these Proceedings for 1900, p. 451. Iowa and Arkansas. All of the above references pertain, I believe, to a single widespread race of P. albolabris, varying in size and color almost as much as the Eastern form of the species, but in a broad view distinguishable from the latter by one, several or all of the following characters : The shell is thinner, more depressed and more glossy; the spiral lines and other minute sculpture are weaker; the lip is narrower, rounded rather than if at, with a weaker less angular rib within; the low basal tooth is fre- quently more distinctly defined. Distribution, west of the Mississippi from southern Minnesota to Arkansas, and eastward in the South to Jackson county, in northern Alabama. In the North, P. albolabris replaces alleni east of the Mississippi river, in Illinois. I have seen typical P. albolabris from west of the Missis- sippi only from Winfield, Henry county, in southeastern Iowa, where it coexists with var. alleni. Owing to the frequent cutting of "ox- bows" by the Mississippi, and the consequent transfer of islands from one to the other side of the stream, even that great river is no bar to the distribution of snails inhabiting lowland forests; and somewhere a,long the immediate vicinity of the Mississippi the areas of albolabris and alleni probably overlap, with perhaps a belt of undifferentiated intermediate forms. 198 PROCEEDINGS OF THE ACADEMY OF [Feb., Specimens of P. a. allcni from Albert Lea, Minn., the most northern point from which I have seen the variety, those from Winfield, Henry eounty, in southeastern and Des IMoines in central Iowa, are yellow or whitish, small, averaging 23 mm. in diameter; 70 per cent, of those seen measuring from 22 to 24 mm., the extreme of size being 21 and 26 mm., these extremes represented by very few specimens, I have below plotted the curve obtained from measuring the diameters of a lot of 50 specimens from Des IMoines, la., collected by Mr. T. Van Hyning. The number of individuals is prol^ably sufficient to afford a perfectly normal curve for the particular place these were obtained; but the specimens from Albert Lea, Minn., and Winfield, la., though few in number, are so similar that I would not expect much difference in their curves.* 40% 35X 30^ / / / \ 20Z »' ' 'x 1 1 1 / \ \ \ C •.^ --:-- / / \ U a e> 21m-m 22. 23. 24. 26. 26. 27. 28. 29. 30. Fig. 1, a-b, diameter variation curve of .50 shells of Polygyra albolabris aJleni from Des Moines, la. ; c-d, 18 shells from Seligman, southwest- ern Missouri. This contrasts strongly with the conditions in Avestern Arkansas. '' This curve is remarkably symmetrical, the average diameter of the lot. obtained by dividing the sum of their diameters by the number of specimens, prac- tically coinciding with the diameter of the greatest number of individuals, repre- sented by the mode or highest point of the curve. Variation in size is about equal toward both minimum and maximum. Of course it is understood that exact measurements of the shells would efTace the angles of the "curve," espe- cially'near the mode. I neglected fractions smaller than ^mm., as in organisms of this size greater exactness would be mere pedantry. A widely different condition is seen in the other curve, in which the average diameter is less than that of the greatest number of specimens. It must be constantly borne in mind that tlie curve of the Seligman lot is probably widely different from what a series of 50 or 100 specimens from that locality w^ould produce; but still the absence of individuals between 22 and 25 mm. diameter, together with the fact that other localities in western Arkansas show a large and a small form, leads me to suspect that a sufficient series would form a bimodal curve, the minor mode remote from the other, and probably near the 21 mm. lien. 1903.] NATURAL SCIENCES OF PHILADELrHIA. 199 southwestern Misso\iri and the Indian Territory, where in many locaUties there is a very much wider variation in absolute size. Thu.s, in a series of 18 shells from Sehgman, Barry covmtj'', southwestern Mis- souri, the variation is from 20^ to 30^ mm.; the smallest being smaller than the extreme of the Des ^loines lot of 50 shells, while about 70 per cent, of the shells are larger than the largest from Des Moines. Of course no normal cur\-c can lie plotted from so small a series as 18 specimens; but in view of similar variation in other smaller lots, the general features of the curve are indicative of the wide range of individual variation. Tlie specimens in the present collection from Missouri and Arkansas are as follows. There are not enough from an}^ one place to make their variations of much significance, but I give the data for what they are worth : De Soto, Jefferson county. Mo. A single thin, small specimen of typical alleni, diam. 23 mm. Seligman, Barry count}', southwestern Missouri. See notes above. Many of the specimens are large, 28 to 30 j mm. diam. Chester, Crawford county, Ark. Three specimens measuring 30, 26 and 23 mm. diam. The smaller ones thin, like maritima, the larger one strong. Hardy, Sharp comity, northeastern Arkansas. Four specimens, measuring 21, 22, 24 and 26 mm. diam. Hot Springs, Garland count}', Ark. A single specimen, diam. 29^ mm., less depressed than usual. Rich ^Mountain, Polk county. Ark. Rather heav}', large specimens, diam. 30 mm or more. Carrion Crow Mountain, Pope county. Ark. Diam. 24 to 27-j mm.; rather light, with the usual narrow peristome. Tushkahoma, Choctaw Nation, I. T. Four large specimens only, measuring 28^, 28^, 29A. 3H nun. diam. The spire is a little higher than usual, and there are nearly ok whorls. The peristome is typical of alleni. Standley, Choctaw Nation. A single shell similar to the above, diam. 31 mm. In Kansas P. alleni occurs in Shawnee count}', and at Quenemo, Osage county. The specimens resemble those from Iowa. None have been seen from southwestern Arkansas or Texas. East of the Mississippi, specimens of typical alleni, resembling those of Missouri, are in the collection of the Academy from near Decherd, Franklin count}', in niiddlc Tennessee, near the Alabama boundary. 200 I'UOCEEDINGS OF THE ACADEMY OF [Feb., received from Mr. A. G. Wetherby, and from Chattanooga, collected by Mr. S. N. Rhoads. At the latter place it coexists with a heavy form of albolahris, though the two do not occur together. Polygyra albolabris fuscolabris nov. Finally, near Woodville, Jackson county, in northern Alabama, a form was collected several years ago, by J\Ir. H. E. Sargent, which seems to be the culmination of the alleni type, with some special features which make it advisable to treat it as a local variety, under the new name P. albolahris fuscolabris. The shell is very large, alt. 19, diam. 34 mm., to alt. 21, diam 38 mm.; depressed, glossy and finely striate, as in alleni with the aperture more oblique and the lip nar- rower than in albolabris of the same size. The basal lip bears a low, flat-topped, wide tooth near the columella, and the outer lip is tinted fieshy-brown (fading in cabinet specimens). Whorls 5^. This particular form is yet known from Mr. Sargent's collections only. A large, heavy form of P. albolabris, probably referable to var. major, also occurs near Woodville. P. a. fuscolabris bears such a rela- tion to alleni as major to albolabris, or var. normalis to typical P. anclrewsce. Polygyra zaleta Binn. (ccokln Binn.). Seligman, Barry county, Mo. Small specimens, 20-22 mm. diam. It is in the collection of the Academy from two other localities in Arkansas : Eureka Springs (F. A. Sampson) and ^tlabelvale (C. W. Johnson). Polygyra indianorum tPilsbiy). P. divesta indianorum Pils., Nautilus, XIII, p. 39 (August, 1S99). Ferriss, Nautilus, XIV, p. 28 (July, 1900). Tushkahoma, Choctaw Nation, I. T. Five specimens 25^ to 28 mm. diam.; also Standley,* I. T. Further experience with this form con- vinces me that it is quite distinct from P. divesta; a course prompted by Mr. Ferriss. As he has collected and examined more specimens than any one else, T am the more disposed to depend upon his judg- ment. Polygyra divesta (Old.). Pilsbry, Nautilus, XIII, p. 38 (1899). Chadwick, Christian county, and Seligman, Barry county. Mo. In Ai'kansas at Chester, Crawford county; Petit Jean, Yell county; Hot Springs, Garland county; Carrion Crow Mountain, Pope county. There is but little variation in the specimens from Ai'kansas, all of which are typical. At Seligman, in southwestern Missouri, the size 1903.] NATURAL SCIENCES OF PHILADELPHIA. 201 varies, tlie average being smaller than Arkansas shells. The smallest specimens measure 15 mm. diam., whorls -ih; the largest 18 mm. with 4f whorls. At Chadwick, Mo., the smallest specimens I have seen were taken, varying from 14^ to 17^ mm. diam. ^olygyra binneyana Pil?- Pilsbry, NautUus, XIII, p. 38 (August, 1899); Proc. A. X. S. Phila., 1900 p. 451; Ferriss, Nautilus, XIV, p. 28 (July, 1900). Poteau* Choctaw Nation, I. T.; Rich Moimtain.=^= Polk county, Ark. Polygyra appressa perigrapta Pils- Chester, Crawford county, and Petit Jean, Yell county. Ark. vSpeci- inens typical but ratlier small, diam. 18-19Vmm. A specimen was also sent from Gulf port, Miss. Polygyra thyroides (Say). Cf. Proc. A. N. S. Phila.. 1900, p. 452. Seligman, Barry county. Mo. Small specimens of the globose bucculenta form; diam. 15^ to 18 mm; umbilicus generally nearly covered, as in P. clausa, but sometimes entirely closed. The speci- mens have a decided resemblance to P. clausa. Two out of eight adults have a small parietal tooth. Antlers, Choctaw Nation, I. T.; Poteau, in the same district, near the Arkansas iDoundary. Small specimens, diam. 18 mm. Carrion Crow Mountain, Pope county, also small. At Hardy, Sharp county, in northeastern Arkansas, two specimens of a small form of thyroides were taken, measuring 17^V and 18 mm. diam., and closely resembling P. clausa, from which they differ, how- ever, in the more "dished" lip, less elevated contour, and slight angu- lation of the front of the last whorl. One of these specimens is almost imperforate. Polygyra clausa (Say). Springfield, Green county. Mo. The specimens are quite typical. This is. so far as I know, the southwestern extreme of the known distribution of the species. Two specimens were taken at Hardy, in northeastern Arkansas. In Missouri and Arkansas P. clausa is sometimes very difficult to distinguish from the small form of P. thyroides, which occurs in that region. This is the more remarkable because the two species in other parts of the country are quite readily distinguishable by the shells. Polygyra pilsbryi Ferriss. PI. IX, figs. 1, 2, 3. Ferriss, Nautilus, XIV, p. 29 (July, 1900). Rich Mountain Station, Polk county. Ark., is thus far the onlv local- 202 • PROCEEDINGS OF THE ACADEMY OF [Feb., ity for this species, which in its cuticiilar processes is the most aberrant of the gentts. The original specimens were denuded of the long fila- ments characteristic of the species in a fresh condition, and hence the published description did not mention them. These filaments are long, more or less curved, flexible, and arise in triangtilar laminae from the narrow riblets of the surface. They stand in three principal rows on the body-whorl, one at the periphery, the others above and below it, the upper one ascending the spire midway between sutures. The inner 2h or 8 Avhorls are free from filaments. The base is encircled by two minor and imperfect rows inside the subperipheral one. In all other Stenotremes except P. harbigera (Redf.) the cuticular hairs form a comparatively close pile, and they are arranged in oblique sweeps, or are merely adnate and prostrate appendages trending in the direction of growth-strise. In no other do they form a series of circular, concentric fringes. P. harbigera has a single fringe of similar filaments, usually persisting at the suture only. Polygyra labrosa (J'.land). PI. IX, figs. 4, 5, c. Helix labrosa Bid., Ann. Lye. Nat. Hist, of N. Y., VIT, 430 (1S61). Stenotrema lahrosum (Bid.), JBinney, Man. Amer. Land Shells, 274. Petit Jean Mountain, Yell county, western Arkansas; Seligman, J^arry county, southwestern Missouri. Also Chester,* Hot Springs,* Carrion Crow Mountain, Ark.* Figured for comparison Avith the other Southwestern Stenotremes, the figures published hitherto being in outline and ciuite inadequate for exact comparisons. The Tennessee specimens I have seen are more distinctly striate beneath and much paler than those from Arkansas, the aperture white. From the data now available it looks as though the range of P. labrosa is discontinuous, being interrupted by a wide strip of low country along the Mississippi; but its distribution in Tennessee and Alabama calls for further investigation, as no definite locality in either State is given, either in the books of Mr. Binney or on the labels of specimens from Bland in the Academy collection. Sampson has reported P. labrosa from eleven counties in middle and western Arkansas. Polygyra uncifera Pilsbiy. Pi. IX. figs. 7, 8, ',). 10. P. hirsuta uncifera Pils., Proc. A. N. S. Phila., 1900, p. 453 (September 27, 1900); Ferriss, Nautilus, XIV, p. 30, No. 139c (1900). Polk county, western Arkansas, at Mena and the adjacent Chastat Mountains, and Rich Mountain. The peculiar structure of the peristome of this form certainly indi- cates a distinct species, and not, as I at first thought, a subspecies of 1903.] NATURAL SCIENCES OF PHILADELPHIA. 203 F. hirsuta. Sometimes the parietal lamella is recur\-od at the outer end, hook-like, as in the type (fig. 7), ))ut in most of the specimens from Rich Mountain the recurved part is quite separate from the lamella, standing apart as a short entering denticle (figs. 8, 9, 10). I have observed no other variation of note except in the size and number of whorls, the Rich Mountain series measuring from 6.5 mm. diam. with 4j whorls to 8.2 mm. with 5j whorls. The "fulcrum" is quite long, as in typical hirsuta, and notched above and below. Polygyra hirsuta has not been reported from Arkansas, so far as I know, and evidence is still wanting that it occurs west of the Missis- sippi south of Kansas and Pettis county, Mo., whence it is recorded by Sampson. I regard the specimens reported by Dr. Stearns from the "banks of Yaqui river near Guaymas"' as probably Eastern shells accidentally mixed with those from the locality named. The south- western range of P. hirsuta is apparently much more limited than that of many of the Eastern snails. Poly&yra blandiana n. sp. PI. IX, figs, il, 12, 13. Shell imperforate, depressed, obtusely angular at the periphery, the spire slightly convex, base much more convex, owing to the high position of the peripheral angle. Chestnut-brown; the surface rather glossy and partially dull; nearly smooth, having weak growth-wrinkles above, fainter on the base; without hairs or their scars. Whorls 4§, moderately convex, the last shortly deflexed in front. Aperture re- sembling, in general, that of P. hirsuta, but narrower; the outer end of the parietal lamina is abruptly bent inward, and its crest a little notched. The notch in the basal lip, which seems shallow in a basal view, is seen in front or oblique view to be deep, with a raised border. The outer curve of the lip is rather deeply notched. Fulcrum long, at a right angle with the aperture. Alt. 4, diam. 8 mm. Alt. 3.8, diam. 7 mm. Springfield, Green county, southwestern Misscuu-i, copiously; Selig- man'* Barry county, Mo., 3 specimens. This very distinct race differs from P. hirsuta in the following par- ticulars: The shell is much more depressed and obtusely angular at the periphery; it shows no trace of hairs; the parietal lamella is more curved and stands nearer the basal lip, thereby causing the aperture to be narrower; the notch in the basal lip is larger, etc. In a rather large series of shells, no specimens at all intermediate in characters were found. ^Proc. U. S. Nat. Mus., XVII, 162 (1894). 204 PROCEEDINGS OF THE ACADEMY OF [Feb., Polygyra monodon imperforata Pils. Proc. Acad. N. S. Phila., 1900, p. 4.5.5. Rich Mountain, Pollv coimty. Ark. The ordinaiy, widely distributed P. m. f rater na occurs in southwestern Missouri. Polygyra monodon aliciae (Pils.). Drift debris of the Arkansas river, at Petit Jean Mountain, Yell county, Ark.; Hot Springs, Ark. ; DeKalb,* Ark., and Antlers,* I. T. BULIMULID^. Bulimulus dealbatus (Say). Seligman, Barry county, Mo. Two gray-mottled specimens collected have the spire much longer than in ordinary dealbatus, and the whorls, except the last two, are more strongly striate. The form approache.s var. ragsdalei to some extent. PUPID^. Bifidaria armifera (Say). Seligman, Barry county, ]\Io. Bifidaria contracta (Say). Seligman, Barry county, Mo.; Rich Mountain, Polk county, Ark. The season was not a propitious one for collecting minute snails, hence their scarcity in this collection. AOHATINID^. Cochlicopa lubrica (MiilL). Seligman, Barry county, Mo. OIRCINARIID^. Circinaria concava (Say). Petit Jean,* Yell county, Ark. ZONITID.^. Vitrea hammonis (Strom). Rich Mountain, Ark.; Seligman, Mo. Vitrea indentata (Say). Poteau, Tushkahoma and Antlers, I. T.; Rich Mountain. Ark.; Carrion Crow Mountain, Pope county. Ark.; Seligman, Mo. Section PAR A VITREA Pils. V Pamvitrea Pils , Nautilus, XI, p. 130 (March, 1898). Taxeodonta Pils., Nautilus, XI, p. 1:^2 (March, 189S). Shell depressed or discoidal, perforate or narrowly umbilicate, com- posed of numerous closely coiled whorls, usually grooved radially above; 1903.] NATURAL SCIENCES OF PHILADELPHIA. 205 thin and fragile; internally having obliquely radial laminae, or rows or pairs of teeth, at intervals of a third of a whorl, some or all of them often wanting, especially in old individuals. Genital system without dart sac. Radula with teeth of the central row tricuspid; two or three laterals on each side also tricuspid, the entocone raised high upon the mesocone, of which it forms a lateral spur. Marginal teeth of the usual simple and thorn-like form. Snails of this group have the discoidal shape and closely coiled whorls of the typical (European) section of Vitrea, and they have also a den- tition of the same type, which is remarkable for the peculiar mode of specialization of the lateral teeth. Paravitrea differs from typical Vitrea in the umbilicate axis and the development of teeth. The compact coiling of the numerous whorls sufficiently distinguishes Paravitrea from the Hyalina type of Vitrea, such as V. hammonis, etc. The species of Gastrodonta protected by teeth, have only a single pair near the aperture, constantly added to in front and absorbed behind with growth of the shell ; but in Paravitrea successive sets are formed to be absorbed later. The adult or old individuals of species of Paravitrea usually absorb all the teeth, and form no new ones in the latest stages; or in some cases, as in V. capsella, teeth may be formed in occasional or rare very young individuals, while in the intermediate and later stages of growth none are developed. In a few other species, such as V. clappi, V. sim.psoni and V. placentula, no teeth have yet been observed even in the young. If my interpretation of the facts is correct, such species as V. andrewsoe and multidentata, which commonly possess teeth in adults, are old, relatively unchanged types; while forms toothless at all stages are the most evolved. Paravitrea tJius consists of species and races in various stages of reduction and loss of teeth, but descendants from an ancestral stock which had them. Another modification of the ancestral radially toothed Paravitrea is seen in certain species in which even-edged or serrate radial ribs or lamellae replace the rows of teeth. The irregular or serrate edge of this lamella in some individuals indicates that it has been formed by coalescence of a row of teeth, the intervals between them becoming filled up, exactly as in the Clausiliidoe of eastern Asia the lunella has been formed by coalescence of a primitive row of palatal plicifi.'* In this phylum, too, the armature has been lost in some species and in some individuals of species normally toothed; and by acceleration, ex- ® See these Proceedings for 1901, LIII, p. 638. 206 PROCEEDINGS OF THE ACADEMY OF [Feb., ceedingly rare individuals of species which normally have rows of independent teeth acquire the radial barriers of forms more advanced in the scale of evolution. This is exemplified by V. muUidentata, dis- cussed in a previous paper/ and V. andrewsce, mentioned below. All this merely goes to show that Paravitrea is in a condition of rapid Teadaptation. Mutation takes place chiefly in two definite directions : (1) Toward progressively earlier loss of internal armature, which cul- minates in toothless species, and (2) the jormation of transverse barriers by coalescence of the teeth. Under these circumstances, the path of the systematist is an excep- tionally thorny one, although to the evolutionist the group is full of interesting suggestions. No rigorous definitions are in order; and the distinction between '^species" and "subspecies" — always more or less indefinite — becomes largely a subjective one when large series of specimens are studied. Only a series of individuals representing the whole life cycle, from youth to old age, can give one an adequate idea of the characteristics of the special form from any given locality. Through the kindness of Mr. Ferriss and others, I have been able to study large series of most of the species, in addition to the collection of the Academy. Aside from Mr. Binney's volumes, which contain merely descriptions of part of the species discussed below, the strongest paper dealing with them is Dr. V. Sterki's Notes on Zojiitidce? The fact that Zonites andrewsi and the Eastern form commonly known as significans have few teeth or none when adult is set forth, and the author shrewdly suggests that these two forms may be the young of Z. placentulus and capsella respectively. Dr. Sterki worked from small series of specimens from a few localities, and with much more ample material it is easy now to criticise his solution of the difficult problem; but he saw fur- ther into the relationships of these forms than any wi-iter up to that time. My former course in separating the species into two sections, one of which {Taxeodonta) was placed under Gastrodonta, was a step backward. No such arrangement would have been entertained had I examined the dentition. In Gastrodonta the lateral teeth are more numerous than in Vitrea or Paravitrea, and conspicuously different in shape. ' These Proceedings for 1900, p. 145. * Nautilus, VII, pp. 13-17, June, 1893. It may be as well to note here that the Gastrodonta described by him under No. 3 (p. 14) is what I described as G. collisella. His No. 6 (p. 15) is apparently G. coelaxis. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 207 Key to average or normal specimens of Paravitrea. I. — Surface closely and regularly rib-striate or grooved, at least above; umbilicus minute. a. — ^Diam. 5-6.5 mm. ; spire flat or slightly convex, the suture deep ; 6^ convex, narrow whorls, subregularly and deeply grooved radially, above and below. Umbilicus minute. No in- ternal teeth, V, clappi. a'. — Diam. 2.5 to 3.5 mm.; very closely rib-striate above; whorls 5^-6, closely coiled, the spire but slightly convex. b. — One to three radial rows of 5 or 6 teeth each, visible through the base, T^. multidentata . h' . — One to three radial curved barriers, visible through the base; surface not striated spirally; width of umbilicus about one-tenth that of the shell, . . V. lamellidens. 6".— With barriers like the preceding species, or none; surface finely striate, the stria? decussated by very minute spirals; umbilicus wider, its diameter contained about 5^ times in that of the shell, V. walkeri. II. — Surface sculptured with spaced, unequal radial grooves, more ' conspicuous above. a. — Diam. of adults 5 to 6 mm. ; toothless or with teeth in pairs. b. — Periphery median; aperture usually crcscentic. Eastern forms. c. — Shell toothless, V. capsella. c' . — Shell with internal pairs of teeth, . Y . c. lacteodens. b' . — Western forms (Missouri, Arkansas, Indian Territory), with the aperture subtriangular. c. — -Young usually toothed; adults dome-shaped, with subbasal periphery V.significans. c'. — Not toothed; adults discoidal, . . . V.simpsoni. a'. — Diam. of adults 7 to 7^, mm. ; toothless or with 1 to 5 teeth in each row. b. — No teeth at any stage of growth; wJiorls 7 to 7h, V. placentula. b' . — Teeth developed in young shells, persisting or absent in adults; shell very fragile; luhorls S to Sh, V. andrewsce. Titrea clappi (Pils.). PL X, figs, l, la, lb, 8, 6a. Gastrodonta clappi Pils., Nautilus, XII, p. 86 (December, 1898); XV, p. 37, PI. 2, figs. 8, 9 (August, 1901); Proc. Acad. Nat. Sci. Phila., 1900, p. 144. Ferriss, Nautilus, XII, p. 99; XV, p. .58. Range : Great Smoky Mountains, along the Tennessee-North Caro- lina boundary range from Thunderhead and Blockhouse Mountains to the Big Pigeon river. The type locality is Miry Ridge. The type (PI. X, figs. 1, la, lb), with a diameter of 5,5 ram., has a minute umbilicus, about .2mm. wide, and the spire of 7 whorls is convex. In all other specimens seen the 208 PROCEEDINGS OF THE ACADEMY OF [Feb.. spire is lower, either flat or but slightly convex; and with one excep- tion, all have the same minute umbilicus. In one specimen, from Clingman Dome, No. 2,490 of the collection of Mr. George H. Clapp (PI. X, figs. 8, 8a), the umbilicvis is decidedly larger, .6 mm. wide, the shell having a diameter of 6.2 mm., with 6^ whorls. Whether this is an individual variation or the representative of a race of more widely imibilicate shells, remains to be determined. In 1900 Mr. Ferriss traced this beautiful and excessively rare species far to the northeast of the original locality, taking specimens at Mt. LeConte, and at Indian Gap and Newfound Gap on the Big Pigeon river. Vitrea lamellidens ( Pils- 1. PL X, figs. 2, ;% 3a, -.ib. Gastrodonta lamellidens Pils., Nautilus, XI. p. 1.34 (April. 1S98); Proc. Acad. Nat. Sci. Pliila., 1900, p. 145. Ferriss, Nautilus, XII, p. 99; XIV, pp. 52, 58. Walker and Pilsbry, Proc. Acad. Nat. Sci. Phila., 1902, p. 437. Range : Great Smoky Mountains, along the Tennessee-North Carolina boundary, from the Little Tennessee river to Thunderhead Mountain (the type locality), and eastward in Graham county, N. C. ; also in the Black Mountain range at Bluff Mountain and Pinnacle of the Blue Ridge. Similar to V . multidentata, but armed within with curved, obliquely radial barriers instead of rows of teeth, and having a narrower imibili- cus, its width contained about 10 times in that of the shell, while in multidentata it is contained only about 6 times. In about 150 individuals collected in 1899 and previoasly, which I ex- amined, all had from one to three lamellae; but a very large shell, diam.. 3.8 mm., found by Mr. Ferriss on Thunderhead in 1900, has no lamellae whatever — a feature of senility. Even in the youngest individuals I have seen, such as PI. X, fig, 2, diam. 1.4 mm., there are two or three barriers, though one would expect rows of teeth in so early a stage. Not one T'. multidentata has yet been found with lamellidens, which lives mainly in the moist heights, where the mountains lift their heads into the clouds. We did hot find it in the "coves" below. Vitrea multidentata (Binney). PI. X, figs. 6, 6a. W. G. Binney, Man. Amer. Land Shells, p. 183 (1885). The southernmost locality is Talassee Ford of the Little Tennessee river, whence Mr. Ferriss reports it.^ Figured for comparison with the preceding and following species. With ordinary specimens of midlidentata there sometimes occur shells in which radial barriers similar to those of V. lamellidens replace the 9 Nautilus, XIV, 58. ~~~ 1903.] XATUKAL SCIENCES OF PHILADELPHIA. 209 teeth of tlie normal form. Such specimens are known from West Granby, Hartford comity, Conn.; Garrettsville, 0. (PL X, fig. 7); Greenwich, N. Y. (Acad. Coll.); Deering, N. H.; Litchfield, N. Y. (Coll. G. H. Clapp) ; Ottawa, Can. (Coll. Bryant Walker),!" and Ithaca, N. Y. (Coll. H. E. Sargent). In the adult specimens I have seen, like thatfigured (diam. 3 mm.), the umbilicus is about as wide as in multi- dentata. In those examined by Mr. Clapp it is narrower, as in lamelli- dens; but his shells are not full grown, measuring only 2 to 2.5 mm. in diameter. From the occurrence of this form as rare single speci- mens, always with typical multidentata, I am still disposed to look on the specimens as accelerated individuals, sporadically occurring — the pioneers of a new race, further advanced in evolution than multi- dentata. If my view is correct, this race of the future will be parallel to lamellidens, which has in the past undergone a like transformation. A view apparently much simpler is held by Mr. Clapp, who regards the Northern specimens with continuous barriers as really V. lamelli- dens, which in this view has an extensive range north to Canada. Vitrea walkeri (Pilsbry). PI. X, figs. 4, la, 5. Gastrodonta xvalheri Pils., Proc. Acad. Nat. Sci. Phila., 1900, p 146 Ferriss Nautilis, XIV, pp. 52, 58. Range : Great Smoky Mountains, in Graham county, N. C, and Mon- roe county, Tenn. This form is about the size of V. multidentata, ordinarily measurino- about 2.9 mm., with 5^ to 5f whorls, and an umbilicus contained about 5^ times in the diameter of the shell, much as in midtidentata. It is either toothless or has radial barriers of the lamellidens type, but the wider umbilicus and minutely decussate surface readily sep- arate the species from lamellidens. No specimens have turned up since our expedition of 1899, and it remains one of the rarest of land snails. Figs. 4, 4a represent a shell from Tuskeegee Mountain; fig. 5 a specimen from Talassee Ford of the Little Tennessee river. Vitrea andrewsae (W. G. Binney). PI. XI, figs. 9, 9a, 10, 11, lla, 116. Zonites andrewsi W. G. B., Ann. N. Y. Acad. Sci., I, p. 359, PI 15 fig D (1S79); [First] Supplement to Terr. Moll., V, p. 144, PI. 2, fig. D (reprint from Ann. N. Y. Acad.) ; Man. Amer. Land Sliells, p. 228, fig 251 Sterki Nautilus, VII, p. 16 (1893). > B o . oierKi, Gastrodonta andrewsce W. G. B., Walker and Pilsbrj^ Moll, of Mt. Mitchell Region, N. C, in Proc. Acad. Nat. Sci. Phila., 1902, p. 437. Range: Roan Mountain (type locality) to Paint Rock on the French Broad river, and eastward to the Black Mountains, N. C. '-° Ottawa Xaturalist, XIV, p. 94, 1900. A single specimen of " lamellidens " with multidentata, collected by Gilbert Heron, is recorded by Mr. Walker 14 210 PROCEEDINGS OF THE ACADEMY OF [Feb., Shell similar to V. placentula m general form and size, but the whorls increase more slowly, the last being narrower ; thin and fragile, often sprinkled with buff dots. Teeth generally present, 3 to 5 being arranged in a radial row, or with several such rows within the basal wall, though the number of teeth may vary down to in the adult stage, or less frequently at any stage of growth. Whorls S to Sh in adults, which measure 7 to 8 mm. diam. In some localities, as around Roan Mountain, teeth are almost invari- ably developed. In some other localities they are rare in adult shells, and developed in the yomig ones only. Such toothless individuals are separable from V. placentula by their slightly more fragile structure and greater number of narrower whorls. In a multitude of shells examined from over a dozen localities, I found one young specimen with a diameter of 4 mm. in which there was a single radial barrier, such as occurs in T'. hnnellidens Pils." The figures are from topotypes. Vitrea placentula (Shuttle worth). PI. XI, Pgs. i, la, 16. Zon. placentuhis Shuttl., Sterki, Nautilus, VII, p. 17 (1S93). Vitrea capsella 'placentula Shuttl., Pils.. Proc. Acad. Nat. Sci. Phila., 1900, p. 140. This species, if such it be, differs from V . capsella merel}' in the larger size, and in having an additional whorl. Adult shells measure 7 mm. diam. , and have 7 or 7^ whorls. I have never seen a specimen with teeth, although a good many shells, adult and young, from the Great Smoky Mountains and elsewhere have passed under my lens; but when very young individuals of placentula are found, I think they will prove to have pairs of teeth, at least in rare eases, as in capsella. It is a trifle more solid than V. anclrewsw (W. G. B.), and in specimens of the same size the latter has a whorl more. The range of V. placentula lies to the southwest of that of V. andrewsa. It was not taken by Ferriss and Walker in the valley of the French Broad river, nor in the Black mountains. Sterki's suggestion that andrewsai is the young of placen- tulus is not borne out by the facts now known. Over a large part of the range of placentula, V. andmvsce does not occur. The figured specimen is from Philadelphia, Toudon county, Tenn. Vitrea capsella (Old.). PI. XI. figs. 2, 3, 4, 4a, ib. Zonites capsella Gld., W. G. Binney, Man. Amer. Land Shells, p. 221. The type locality of this species is ''Tennessee." Two specimens "^ This specimen is No. 1G,963 of Mr. Bryant Walker's collection, taken by him at Paint Rock, N. C, on the soutli side of the French Broad river, near tlie Ten- nessee boundary. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 211 SO labelled, presumably from the original lot, are in the collection of the Academy, given by Gould. The species differs from V. significans chiefly in the different contour of the adult shell; from V . placentula in the smaller size of the adults. The shell measures from 5 to nearly 6 mm. in diam., and has 6 to 7 whorls. Rarely, a young indi\ddual may be found with a pair or two of internal teeth (PL XI, fig. 3, diam. 2 mm.), recalling the ancestral den- tate stock; but this stage is now passed through at an early age, or entirely skipped ; and in the series I have examined, from southwestern Virginia, Kentucky, Tennessee and Alabama, toothed individuals are very unusual. V. capsella is not known from west of the Mississippi depression, being represented there by the very closely related T'. significaiis and V. simpsoni. The figures represent specimens from Woodvilie, Ala., collected by ^Ir. H. E. Sargent. Vitrea placentula and significans might, with no great violence, be subordinated to capsella as subspecies. Vitrea capsella lacteodens n. v. PL XI, figs. 5, 5a. Hyalina significans Bid., Harper, Journ. Cincinnati Society of Natxiral History, IV, 1881, p. 258, figs. 2, 2a. Zonites significans Bid., Wetherby, Journ. Cin. Soc. N. H., IV, December, 1881, p. 328, No. 25. W. G. Binney, Man. Amer. Land Shells, p. 228, fig. 250 (exclusive of quotation from Bland, etc.). Sterki, Nautilus, VII, pp. 16, 17 (1893). Gastrodonta significans Bid., Pilsbrv, Moll, of the Great Smoky Mts., Proc. Acad. Nat. Sci. Phila., 1900, p. 147. Vitrea capsella, specimens from Tuskeegee Mountain, Graham county, N. C, Pilsbry, t. c, p. 140. Type specimens from "Ramp Cove," Tuskeegee Mountain, N. C, taken by James H. Ferriss and H. E. Sargent, 1899. The shell is similar to V. capsella, except that most specimens have one to three pairs of tubercular teeth within the last whorl. The sutures are a trifle less impressed, and the striation perceptibly closer. It differs from V. significans in the usual persistence of the teeth in the adult stage, and the median position of the periphery. In fuUy adult significans the periphery is subbasal, and there are no teeth. Alt. 2.6, diam. 5 mm.; whorls 6^. This particular race occurs, so far as I know, only in the mountains of southwestern North Carohna. It is not separable from T". capsella by any hard-and-fast character, but merely by the persistence of the pairs of teeth in most adult shells of any given lot. Thus in the type lot, collected by Mr. Sargent on Tuskeegee Mountain, Graham county, N. C, one adult out of thirteen before me is quite toothless, and by 212 PROCEEDINGS OF THE ACADEMY OF [Feb., itself would be called capsella. Some others have only one or two teeth remaining. It is the general character of the specimens from any one place, not the particular condition of each individual, that must be considered. Harper, Wetherby, W. G. Binney and Sterki have mistaken this race for the Western V. significans, and I formerly followed this erroneous identification. Vitrea significans (Bid.)- PI- XI, figs, 7, 7a, lb, s, sa. Helix significans Bid., Amer. Journ. of Conch., II, p. 372, PL 21, fig. 9 (not good). Zonites significans Bid., Binney, op. plur., exclusive of East Tennessee form. Range: Originally described from Fort Gibson, Indian Territory, this species has been taken to my knowledge in Arkansas, at Mabelvale (C. W. Johnson^^), and in southwestern Missouri, at Seligman, Barry county (Ferriss, 1901). As Bland stated in his original description, the young shells are sometimes provided with one or two pairs of tubercular teeth within, visible through the base of the shell as white spots. The full-grown shells are toothless, more or less dome-shaped, the periphery being situated below the middle of the last whorl, the circumference of which is flattened and sloping. The base is very concave in the middle. This gives the shell a peculiar and uniisual contour. Half- grown and young individuals are normal in shape, and very similar to V. capsella and its variety lacteodens, from which, indeed, it would be almost impossible to separate them except by the locality. The spire, seen from above, is about the same in capsella, placentula and signifi- cans. Figs. 7, 7a, 7b represent fully adult specimens of the typical form from Seligman, Barry county. Mo., collected by Mr. Ferriss; figs. 8, 8a are immature sheUs from the same place and lot. Vitrea simpsoni (Pilsbry). PL XI, figs. G, 6a, 6b. Zonites capsella Gld., Simpson, Proc. U. S. Nat. Mus. for 1888, p. 452. Zonites simpsoni Pils., These Proceedings for 1889, p. 412, PI. 12, figs. 8-10. Vitrea simpsoni Pils., These Proceedings for 1900, p. 456; Ferriss, Nautilus, XIV, pp. 30, 31. Range : Western Arkansas and Indian Territory ; Mena and Hatton's Gap, Polk county, in western, Morris Ferry, Little River county, southwestern Arkansas (J. H. Ferriss); Limestone Gap, I. T. (C. T. Simpson). This species is more depressed than significans or capsella, with the last whorl decidedly wider, when viewed from above. It has not j^et '^ Reported in the Catalogue of Land Shells, 1898, p. 26, under T'. simpsoni, No. 275. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 213 been found with teeth, but so far only smaU numbers have been takea. The type is figured. Zonitoides minusculus (Binn.). Sehgman, Barry county, Mo. Zonitoides arboreus (Say). Poteau, Tushkahoma, and Antlers, I. T.; Seligman, Mo. Gastrodonta demissa (Binney). Carrion Crow Mountain, Pope county, Ark. Twenty-one specimens, most of them narrowly perforate, with a very heavy internal callus; four have a low lamella, and one is imperforate. They therefore imite characters of the varieties lamellata and hrittsi, hke those from other places in their same general region. Cf. these Proceedings for 1900, p. 456. At Hot Springs, Garland county, Ark., G. demissa hrittsi was taken, a few of the specimens having an internal lamina. Various forms of demissa were taken at Poteau* and Tushkahoma,* I. T., and Rich Mountain,* Ark., reported by Mr. Ferriss. ENDODONTID^. Pyramidula solitaria (Say). Reported from Rich Mountain, Polk county. Ark., by Mr. Ferriss. Pyramidula alternata (Say). Typical specimens at Petit Jean, Yell county, Ark., and SeHgmar., Mo., and Standley, Choctaw Nation, I. T. Also reported by Ferriss from Rich Mountain and Hot Springs, Ark. Pyramidula perspectiva (Say). Reported by Ferriss from Seligman, Mo.; Petit Jean and Rich Mountain, Ark. Helicodiscus lineatus (Say). Seligman, Barry county. Mo. SUCCINEID^. Succinea avara Say. Seligman, Mo. ; Rich Mountain, .^k. LIMN^ID^. Limnsea columella Say. Winding Stair Mountain, I. T. 214 PROCEEDINGS OF THE ACADEMY OF [Feb., EXPLAXATIOX OF PlaTES IX, X, XI. Plate IX, Figs. 1, 2, 3. — Polygyra 'pilshryi Ferriss. Topotype. Figs. 4, 5, 6. — Polygyra hlandiana Pils. {Helix labrosa Bland, not Wood). Type. Petit Jean Mountains, Yell county, Ark. Fig. 7. — -Polygyra uncifera Pils. Cotype. Mena, Polk county, Ark. Figs. 8, 9, 10. — Polygyra uncifera Pils. Rich Mountain, Polk county, Ark. Figs. 11, 12, 13. — Polygyra hlandiana Pils. Cotype. Springfield, Green county. Mo. Plate X, Figs. 1, la, \h. — Vitrea clappi Pils. Three views of the tj^pe. Miry- Ridge, Great Smoky Mountains. Fig. 2. — Vitrea lamellidens Pils. Young individual from Thimderhead, diam. 1.4 mm. No. 77,752 A. N. S. P. Figs. 3, 3a, 36. — V. lamellidens. Thunderhead. Diam. 3.5 mm. No. 77,752 A. N. S. P. Figs. 4, 4a. — F. walkeri Pils. Front and basal views of the type. Fig. 5. — Vitrea walkeri Pils. Base of shell from Talassee Ford. No. 77,703 A. N. S. P. Fig. 6, 6a. — Vitrea midtidentata Binn. Basal and front views of a typical specimen from West Granby, Hartford county, Conn. No. 57,104 A. N. S. P. Fig. 7. — Vitrea multidentata Binn. Base of a specimen with the teeth united into an even barrier. Garrettsville, O. No. 66,858 A. N. S. P. Figs. 8, 8a. — Vitrea clappi Pils. Openlv umbilicate specimen from Cling- man Dome. No. 2,490 Coll. G. H. Clapp. Plate XI, Figs. 1, lo, \h. — Vitrea placentula Shuttl. Philadelphia, Loudon, county, Tenn. No. 5,490 A. N. S. P. Fig. 2. — Vitrea capsella Gld. Half-grown shell, diam. 3 mm. Wood- viUe, Ala. No. 68,820 A. N. S. P. Fig. 3. — Vitrea capsella Gld. Young specimen with two pairs of internal teeth, diam. 2 mm. WoodvUle, Ala. No. 68,820. Figs. 4, 4a, 4b. — Vitrea capsella Gld. Adult from Woodville, Ala., diam. 4.5 mm. No. 68,820 A. N. S. P. Figs. 5. 5a. — Vitrea capsella lacteodens Pils. Tuskeegee Mountain, Graham county, N. C, No. 77,798 A. N. S. P. Figs. 6, 6a, 66. — Vitrea simpsoni Pils. Type. Limestone Gap, I. T. No. 61,676 A. N. S. P. Figs. 7, 7a, 76. — Vitrea significans Bid. A fuUy adult specimen from Seligman, southwestern Missouri. No. 81,446 A. N. S. P. Figs. 8, 8a. — Vitrea significans Bid. Young shell with internal teeth, diam. 3.8 mm. Same locality. Figs. 9, 9a. — Vitrea andrewsce W. G. B. Half-grown shell from Roan Mountain, N. C, diam. 4.5 mm. Fig. 10. — Vitrea andrewsce W. G. B. Younger sheU, diam. 3 mm. Figs. 11, llo, 116. — Vitrea andrewsce W. G. B. Adult, Roan Mountain, N. C. No. 67,577 A. N. S. P. This and figs. 9, 9a and 10 are from specimens of the original lot, collected by Mrs. George Andrews. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 215 A REVISION OF THE NORTH AMERICAN ANTS OF THE GENUS LEPTOTHORAX Mayr.' BY WILLIAM MORTON WHEELER. The diminutive species of the cosniopohtan genus Leptothorax are among the most interesting though least conspicuous of our ants. No other group of Formicidse appears to present such diversity of habits, while at the same time adhering so closely to certain rather definite generic peculiarities. The species all form small colonies, often of not more than twenty-five to fifty individuals, and occupy small cavities in the soil, or between stones, or in the tissues of plants. These cavities are either of their own excavation, or found ready to their use in the form of accidental openings or the burrows and gal- leries of larvas, other ants, etc. Favorite nesting-places are the abandoned woody galls of the Cynipidae, like the galls of Holcaspis cinerosus on the live-oaks of central Texas, the galls of Diptera, like Trypeta soUdaginis on the golden-rod, etc. Even hollow nuts on the ground under the trees are sometimes tenanted by species of Leptothorax. Brief notes on the nesting habits, so far as these are known, are appended to the descriptions of the different species enum- erated below. The small size and obscure location of the Leptothorax nests, which form a remarkable contrast with the teeming, conspicuous formicaries of other ants like Formica rufa, F. exsectoidcs, Pogonomyrmex barbatus and Ischnomyrmex CockereUi, will readily account for our rather limited knowledge of the North American species. Then, too, none of our Leptothorax are really common, except in certain circumscribed localities, so that the discoA-ery of the species is more often a matter of accident than of deliberate search, even when one is out looking for ants and nothing else. Single workers are found running about on the ground or on the trunks and branches of trees in search of sweet exudations, small insects or the remains of large insects that have been rejected by si^iders, birds, etc. The nests are most readily found by following up such single workers, often a tedious and time-con- suming task, as these insects will sometimes run about for an hour or 1 Contributions from the Zoological Laboratory of the University of Texas, No. 48. 216 PROCEEDINGS OF THE ACADEMY OF [Feb., more in search of food before returning to the nest and reveaUng its hidden entrance, a tiny hole like a pin-prick in the soil or bark. None of the species are known to attend aphides, and the nests very rarely or never contain guests or synoeketes of any description. The Leptothorax themselves, however, sometimes live as guests in the nests of larger ants. Thus L. Emersoni is always found as a guest in the nests of Myrmica brevinodis, and L. acervorum of Europe and its American variety convivialis also exhibit a decided tendency toward xenobiosis. The typical L. curvispinosus appears to act as the slave of Tomognathus americanus, an extremely rare ant, which is probably similar in habits to its European congener, T. suhhwis. JMost of the species of Leptothorax are very timid, and many of them readily "feign death" when roughly handled. Nevertheless they are often extremely hostile and vindictive toward other ants, especially toward ants of their own species from strange nests. For our first insight into the habits of Leptothorax we are indebted to Forel, who recorded his observations in the charming Fourrnis de la Suisse (pp. 339-341). The more important of these observations are given in the following translation : ''April 17, 1868, I found in the bark of a pine-tree a very small colony of L. tuber o-affiiiis, consisting of a fertile female, about a dozen workers and some eggs. I lost four of the workers during the capture and broke two of the legs of the female. I subsequent!}' placed this little family in a pasteboard box with a glass cover. It prospered; the female got on well with her four remaining legs ; some of the eggs hatched and the larvae were fed. The workers would eat nothing but the honey which I gave them; they were very timid and settled down with the female in the box. They gave little heed to the female, which lived almost like them. I have noticed that it is only the workers of the genera Plagiolepis and Lasius and of certain species of Formica that assiduously court their fertile females. Leptothorax goes to the opposite extreme: the females live almost like the workers, being merely some- what less inclined to work. Huber was wrong, therefore, in general- izing the role of the fertile females of Losiiis, etc. By MsiV 24 the female of my captive formicary had again laid some eggs, and the larvse had grown very large. The workers ate larvne of Lasius that were given to them. June 4 one of the larvse became a worker pupa, but there remained only two large larv?e and the eggs. June 10 there were two pupoe and eight or nine small larvse had hatched from the eggs. Of the latter two were yellow and retained this color, the others were whitish. By June 13 thev had srown. I then 2;ave mv L. 1903] NATURAL SCIENCES OF PHILADELPHIA. 217 iuhero-ajjims a worker pupa of L. Nylanderi, and they took care of it. These ants never attempted to escape when I opened the box. They impressed me by the delicacy of the sense of touch in their antennae, as they felt of little particles with remarkable precision and distin- guished their C[ualities (one of the eggs of their ciueen, a grain of dust, a particle of honey, etc.). June 21 I gave them some pupce of Tetra- morum ccBS'pitum which they killed and ate. On June 25 the pupa of L. Nylanderi had hatched, and the residting worker lived on good terms with the tubero-afflnis, working with them. June 28 I lost several workers through carelessness; there remained only the queen with five worker tuhero-affims and the worker Nylanderi; the small larvae had grown considerably and began to pupate. June 29 one of the older pupae hatched and the other soon followed, so that two worker tubero- affinis were added to the colony. The same day I gave my ants worker pupae of T. cwspitiim. They cared for two or three of the younger ones and killed the others which were about to hatch, or rather allowed them to die through neglect. July 15 three more small larvae made their appearance. July 16 one of the two pnpce of T. cwspitum, which they had continued to foster, hatched and lived thenceforth with these ants of a different genus, on the best of terms. July 18 the second pupa of T. ccespituni followed suit, but this worker was somewhat malformed and died in a week. The first Tetramorium, on the contrary, pros- pered apace; it was larger than any of the Leptothorax workers and was conspicuous on account of its activity. It ran about continually in all parts of the box, but kept returning from time to time to the Leptothorax. By July 29 a fresh batch of little tubero-affinis larvae had grown up, and the pupae of the second generation began to hatch. August 16 I placed the seven surviving tubero-affijiis workers and their queen in alcohol, as the colony had suffered considerably during my trip from Zurich to Vaux. It had lived in captivity four months "June 25, 1868, having found a formicary of L. acervorum in the bark of a pine, with a winged female and some female pupae, I captured and preserved it in a box till August 16. Several females hatched in the box. The females of this species are not larger than the workers. I often saw these little females carrying the larvae and pupae about like the workers. Strange to say, nearly all of them lost their wings within two or three days from the time of hatching. I even saw one of them obviously endeavoring to rid herself of her wings by twisting them about. As they w^ere born in a box containing no males, they could not have been fecundated. Hence I cannot conceive why they removed their wings. Can it be that the formicaries of Leptothorax 218 PROCEEDINGS OF THE ACADEMY OF [Feb., are kept up in this way, because the workers cannot retain the fertile females in the nests by force, on account of the small size of the nests and their position on vertical walls, or because the males may not often be present simultaneously with the winged females? Would not these dealated females be fecundated later by males appearing in the same formicary? The fact remains that one often finds in Lcptothorax nests dealated females with small abdomens and apparently not fecundated, together with others obviously fertile. I refrain from de- ciding the question." The questions asked by Forel so long ago still remain unanswered, although it is clear that the colonies are not as a rule renewed and maintained b}^ a retention of the virgin females in the parental nest. My own observations show that the little colonies of these ants are founded by single fertile females, in the very same manner as the huge formicaries of Formica, Camponofus, etc. On several occasions I have found dealated females of Lcptothorax either alone or with a very few eggs, larvae or pupse in isolated oak-galls {e.g., L. obturator q. v.). Moreover, I have never found more than one queen in a nest in any of the species that I have taken, except at the very height of the breeding season (May and early June in Texas, mid- or late summer in the Northern States). Although in such nests I have sometimes seen several dealated females, which probably arose as Forel has described, I am inclined to believe that all of these, except the mother queen, must soon leave the nest and establish colonies of their own. The question naturally suggests itself: Why are the colonies of Lcptothorax so small? I believe that this peculiar condition may be traced, in part at least, to the following causes, either singly or col- lectively: 1. The females are but little larger than the Avorkers (in L. Emcrsoni they are not even larger than the workers) and this means relatively small fecundity. This appears to be the case also in other ants that have females of the same or nearly the same size as the workers (Myrmecina, Stcnamma s. str.; Ponerinae). And recipro- cally, owing to this reduced fecundity, the cjueen cannot be abundantly fed, since she produces but few workers. 2. The workers of Lcpto- thorax are probably short-lived as compared with many other ants. At least one is inclined to believe this from the rather high mortality among these insects in artificial nests. 3. In most species of Lcpto- thorax each colony contains only a single fertile queen.- ^ Other observations on the habits of Lepiothorax will be found in the following works: Adlerz, " Myrmecologiska Studier," II. Svenska Myror och deras Lef- nadsforhollanden, Bihang till K. Sve7iska Vet. Akad. HandL, Bd. XI, No. 18, 1903.] NATURAL SCIENCES OF PHILADELPHIA. 219 The geographical distribution of the North American Leptothorax, though very incompletely knoAvn, is not altogether devoid of interest. So far as it is possible to generalize from existing data, it would seem that the species are rather uniformly distributed over the entire con- tinent, not excepting at least a portion of the Arctic regions. By this I do not mean to say that the same species occur everywhere, or even that the distribution of a particular species is very wide, but that the ant-fauna of any given locahty usually comprises a few species of Leptothorax. This indicates a wide range of adaptability to differences of soil, moisture, temperature, vegetation, etc., within the same genus. The extremes of this adaptation seem to be represented by forms like L. curvispinosus, which inhabits the humid shady woods of the North Atlantic States, and L. Pergandei, w^hich occurs even on the sun- scorched soil of the Trans-Pecos deserts. We have few species in common with Europe, probably only L. acervorum and L. muscorum, both presenting distinct American varie- ties or subspecies analogous to and occurring over the same territory as the American forms of Formica fiisca, rufa and sanguinea and Myr- mica rubra. All of these forms occur far to the north and to consid- erable altitudes, both in Europe and America, and undoubtedly constitute important elements of an ancient palaearctic ant-fauna.^ At lo.w altitudes and within our territory the forms of L. acervorum and muscorum seem to be confined to the northernmost tier of States. The twenty species of Leptothorax recognized in the present paper as occurring in America north of Mexico are about equally distributed between the two divisions of the genus, which are characterized respec- tively by the workers and females having 11- (the males 12-) jointed antennae, and the workers and females having 12- (the males 13-) jointed antennae. It is an interesting fact that the species with 11- jointed antennae in the workers are mainly confined to the Northern and Eastern States, those with 12-jointed antennae to the Western and Southwestern territory. Exceptions are L. curvispinosus and acer- vorum, which present varieties even in New Mexico (though at consid- erable altitudes !) and L. tricarinatus , which was described from South 1886, and III. Tomognathus sublajvis Mayr, ibid., Bd. XXI, No. 4, 1896; Wheeler, "The Compound and Mixed Xests of American Ants," A7n. Natural., Vol. XXXV, Nos. 414, 415, 417 and 418, 1901, and " Ethological Observa- tions on an American Ant {Leptothorax Emersoni Wheeler)," Arch. j. Psych, u. Neurol., Bd. II, Heft 1 u. 2, 1903 pp. 1-31. ^ L. acervorum var. convivialis (q. v.) has been taken on the summit of Las Vegas Range, N. M., at an altitude of 11,000 feet, by Prof. T. D. A. Cockerell in xenobiosis with another boreal ant, Myrmica brevinodis. L. acervorum var. Kincaidi was described by Pergande from Alaska. 220 PROCEEDINGS OF THE ACADEMY OF [Feb., Dakota. All the new species described in the present paper belong to the group with 12-jointed antennae, and it is probable that many more members of this group remain to be discovered in the West and Southwest. Both groups are represented in ^lexico and Central Amer- ica. To judge from Emery's table of the South American species, those with U-jointed antennae predominate again south of the Equator. Most of these species, however, whether having 11- or 12-jointed antennae, have acute, projecting angles to the pronotum, and are there- fore consigned to a particular subgenus, Goniothorax, by Emery.* The small group comprising the subgenus Dichothorax Emery (possibly monotypic) is confined to the Southern United States. This subgenus resembles the subgenus Temnothorax j\Iayr (including only T. recedens Nyl.) in many respects. It is interesting to note that this form occurs only in Southern Europe. While some of the European Leptothorax (like tuberum and uni- fasciatus) are known to be extremely variable, the North American materials at the disposal of previous writers have not been sufficient to prove the same for any of the species on this side of the Atlantic. Nor am I able to throw as much light as I could wish on the limits of variabihty in our species, although my material certainly shows that some of our species are decidedly unstable. Such are, e.g., L. acer- vorum, curvispinosus , nitens and possibly also Schaumi and fortinodis, especially if the two latter realty represent extreme forms of the same species, as seems to be indicated by the existence of intermediate forms. The genus Leptothorax was established by ^layr in 1855,' on a num- ber of species previously included by Nylander and other myrmecolo- gists in the composite genus Myrmica, a genus which at one time contained practically all the known ants of the subfamily Myrmicinae. Though some of the characters of Leptothorax are not very definite, the genus has nevertheless stood the test of nearly half a century and will probably continue to stand. Like many ant-genera, and genera of other animals also, for that matter, it is recognized not so much by a description of its characters, as by its peculiar and almost unmistak- able habitus. He who has had little experience in handling ants will be liable to confound the workers of Leptothorax with the workers of Pheidole or vice versa, but to the experienced eye even the gait of the *'Studi sulle Formiche della Fauna Neotropica," Bull. Soc. Ent.Ital. Ann., XXVIII, 1896, pp. 26, 27. '• ■' Formicina Austriaca," Yerh. K. K. zool.-hot. Ver. Wien, Bd. 5, lSo.5, pp 431-433. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 221 ants of these two genera differs very decidedly. The following are the leading diagnostic characters of the genus Leptothorax. LEPTOTHORAX Mayr. Worker. — Small, monomorphic. Head longer than broad, and broader than the thorax, ^landibles broad, 4-5-toothed. Maxillary palpi 5-jointed; labial palpi 3-jointed. Clypeus variable in shape, slightly convex or impressed in the middle, its anterior border some- what rounded, entire or sinuately excised in the middle. Frontal carinse almost straight, diverging very little behind. Antennse 11- or 12-jointed, usually with a distinctly 3-jointed club. Frontal area present. Eyes of moderate size, near the middle of the lateral surface of the head. Ocelli occasionally present, especially in ergatoid or subergatoid individuals. Thorax slender, usualh^ somewhat broader in front, at least above, and narrower behind. Promesonotal suture obsolete; mesoepinotal suture present or absent, the thorax at this region either without any constriction, with a faint or a very decided (subgen. Dichothorax) constriction. Epinotum armed with a pair of teeth, or spines of variable development. Petiole with a short pedun- cle in front and surmounted by a node of variable form, its lower an- terior surface armed with a median tooth. Postpetiole nodiform, sometimes more campanulate, unarmed below. Gaster large, broadly elliptical, compressed dorsoventrally, its basal three-fourths formed by the first segment. Sting well developed, at least in many of the species. Legs rather stout, the femora fusiform, somewhat incrassated in the middle, the tibiae tMcker toward their distal ends. Spurs of middle and hind legs simple, not pectinate. Integument very hard. In most species the hairs on the body, and in a few also those on the appendages, are short, erect, clavate and under a high magnification finely crenulate. Our species are yellow, brown, red or black, and the majority of them have the head, thorax and pedicel more or less sculp- tured and in great part opaque. The gaster in all our species is very smooth and shining. Female. — Somewhat larger and more robust than the worker, or of the same size. Antennse of the same number of joints. Eyes and ocelli moderately prominent. Thorax with its sides subparallel or somewhat bulging in the middle. Mesonotum conspicuously flattened. Epi- notal spines shorter and often stouter than in the worker. Basal surface of epinotum nearly horizontal. Petiole and postpetiole like the corresponding segments of the worker, the node of the former often more acute. Gaster like that of the worker, at least its basal two- ooo -^- PROCEEDINGS OF THE ACADExMY OF [Feb. thirds formed by the first segment. Wings milky or yellowish hyaline, with very pale and indistinct veins and stigma. Radial cell sometimes open, sometimes closed. There is a single cubital cell. Transverse vein meeting cubital vein at its bifurcation; internal cubital often indistinct. Discal cell closed. The pilosity of the female is usually less pronounced than that of the worker, though of the same character; the sculpturing is rougher. Male.—Oi the same size as the worker, or but httle larger, and usually darker in color. Head short and as broad or broader than the thorax. Mandibles variable, narrow, truncate and toothless, or den- tate or denticulate. Clypeus somewhat convex. Antenna 12-13- jointed; scape short, funiculus very long, slightly thickened at its distal end to form, in many cases, an indistinct 4-jointed club. Eyes and ocelli large and prominent. Mayriaii furrows of mesonotum very distinct. Epinotum not prolonged backward, with two small swell- ings, rarely with two short teeth, in the place of the worker armature. Petiole more slender and with lower node than in the worker. Post- petiole nodiform or subcampanulate. Gaster rather slender, elongate elliptical, often shghtly flattened dorsoventrally. Legs' slender. Wings as in the female. Hairs on the body and appendages usually much less conspicuous than in the worker, not clavate. I subjoin a table for the identification of the workers of the various Leptothorax species known to occur in America north of Mexico. As the females of only half and the males of less than half of our species are known, it is hardly worth while to construct tables for the identifi- cation of the winged sexes. ^ 6 The following species have been described from Mexico and Central America and m all probability comprise but a very small portion of, the species actually occurring in these countries: f ^ ^^ ciLtudiiy 1. Leptothorax Stolli Forel, Bull. Soc. Vaud. Sci. Nat. (2), XX p 3.5'^ 1894 fm^,'lltl899!75?'-$'^"""'P'-''^"' '^^''' P- '""' ForeX^BfolT'Celtr: Guatemala, summit of crater of the Volcan de Agua (13 000 feet')- Uvino- imdersrones. ' ^^' "'"» 2. L. echinatinodis Forel, Compt. Rend. Soc. Ent. Bek XXX IS^ifi n vlviii ^ ; Biol. Centr. Am., Ill, 1899, p. 5.5. ^" ' ' ^^ "' Rio Janeiro, Brazil. Lives in hollow twigs. The t>T)ical form has not been it°s variS America, but only the foUo^-ing subspecies and possibly Subsp. aculeatinodis Emery, Bull. Soc. Ent. Ital XXYUI 1896 d 60 8 Costa Rica, Jimenez. Also Brazil. ' ' , i ■ ^^. ^ . Var. pungent inodis Emery, Bull. Mus. Zool. Torino XI 1896 p "^ O Mexico, Atoyac en Vera Cruz ; Panama. ' ' ' ' • -• -• 3. L. Pittieri Forel, Biol. Centr. Am., Ill, 1899, p. 56. ^ . Costa Rica. 4. LTristani Emery, Bull. Soc. Ent. Ital., XXVIII, 1896 p 61 xQ Forel, Biol. Centr. Am., Ill, 1899, p. .56 Jimenez, Costa Rica. * 1903.] NATURAL SCIENCES OF PHILADELPHIA. 223 Table for the identification of the workers of Leptothorax. 1. Antennoe 11-jointed, -■ Antennae 12-jointed, ^^ 2. Thorax with faint but distinct mesoepinotal constriction, 3 Thorax without mesoepinotal constriction, ' 3. Postpetiole sculptured, opaque, ^ Postpetiole smooth and shining above, o 4. Tibice and antennal scape witliout clavate hairs, -5 Tibise and antennal scape with short, erect, clavate hairs, L. hirticornis Emery. 5. Hairs on body long and thin, not clavate, muscorum Nyl. var. sordidus var. nov. Hairs on bodv short and clavate, acervorum Mayr. subsp. canadensis Prov., etc. 6. Tibia with clavate hairs, Proranc/im Emery. TibijB without clavate hairs, Emersom Wheeler. 7. Epinotal spines very short, dentiform, 8 Epinotal spines long " 8. Petiole not conspicuouslv larger in profile than the postpetiole, Schaumi Roger. Petiole conspicuouslv larger in profile than the postpetiole, fortinodis Mayr. (and its varieties). 9. Dark-colored, with very long, horizontal epinotal spines, lo7igispinosus Roger. Yellow, sometimes more of less infuscatcd; epinotal spines shorter, curvispinosus Mayr. (and its subspecies and varieties). 10. Without mesoepinotal constriction ; hairs on body clavate, not abundant, 11 With pronounced mesoepinotal constriction, hairs abundant, not clavate (subgenus Dicfu)thorax Emery), 19 11 Head very largely smooth and shining, j-* 'i-' '^'^ Head opaque or subopaque, or smooth only along the median hne or behind, ^"^ 12 Very dark-brown, or black, antennal scape almost reaching the posterior angle of the head, Schmittii sp. nov. Yellow, or somewhat infuscated; scape reaching to | the distance between the eye and the posterior corner of the head, niiens Emery (and its subspecies and variety). 13. Anterior margin of clypeus entire, rounded, .... 14 Anterior margin of clypeus sinuately excised, lb 14. Head and thorax very coarsely reticulate-rugose, . . . texanus sp. nov_. Head and thorax not coarsely reticulate-rugose, l-'5 15. Clypeus produced, with angularly projecting anterior border, tricarinatus Emery. Clypeus not produced, with broadly rounded anterior border, neomexicanus sp. nov. 16. Black or dark-brown species, Y^ Yellow species, ^^ 5. L, petiolatus Forel, Ann. Soc. Ent. Belg., XLV, 1901, p. 129. ^ . Wheeler, Ann. Soc. Ent. Belg., XLV, 1901, p. 201. Cuernavaca, Mexico . "A single nest, consisting of a deiilated queen and about 25 workers in a Tillandsia in parabiosis with Cryptocerus and Cremasii- garter." 224 PROCEEDINGS OF THE ACADEMY OF [Feb., 17. Petiole slender, seen from above three times as long as broad, node A-erj;- low and rounded, obturator sp. nov. Petiole only U times as long as broad; node higher and shorter, nevadensis sp. nov. IS. Head opaque throughout, petiolar node round in profile ; length 1.5-1.75 mm., terrigena sp. nov. Head with a smooth median line, node of petiole somewhat angular; length 2.25, • Andrei Emery. 19. Summit of petiolar node seen from behind impressed or concave, Pergandel Emery, Summit of petiolar node seen from behind convex, rounded, jloridanus Emery. 1. Leptothorax hirticoruis Emery. L. hirticornis Emerv, Zool. Jahrb. Abth. f. Svst , YIII, 1S94, pp. 317 and 319. ^. ' Worker. — Length 2.75 mm, Clypeus hardly impressed in the middle. Antenna? 11-jointed, joints 2-7 of the funiculus shorter than broad, club but relatively little thick- ened. Thorax slender, distinctly impressed at the mesoepinotal suture. Epinotal spines of average size, pointed, strongly com- pressed. Petiole robust, its anterior and posterior dorsal slopes form- ing a slightly obtuse angle; seen from above the sides of the petiole are subparallel. Postpetiole small, almost trapezoidal, a little broader than long. Clypeus somewhat shining. Head, thorax and pedicel opaque, densely foveolate-punctate, the upper surface of the head also finely and rather regularly longitudinally rugose. Hairs very short, strongly clavate, erect, not only covering the body but also the antennal scape and legs. Bright testaceo-ferruginous, gaster and middle of front infuscatcd. Type locality: Washington., D C. (Pergande). Described from a single specimen in the collection of Prof. Emery. 2. Leptothorax muscorum Nylander, var. sordidus var. nov. L. muscorum Emery, Zool. Jahrb. Abth. f . Sj^st., VIII, 1S94, p. 31S. ^ . Worker (PL XII, fig. 2).— Length 2-2.75 mm. Clypeus not impressed in the middle, its anterior border nearly straight, not produced. Antennae 11-jointed; scape reaching to mid- way between the eye and the posterior angle of the head ; first funicular joint as long as joints 2-4 together, terminal joint as long as the two preceding joints of the club. Thorax broad in front, Avith rounded humeri, much narrower behind, with a distinct mesoepinotal con- striction. Epinotal spines of moderate length, directed backward, and slightly outward, about as far apart at their bases as they are long. Petiole from above suboblong, with slightly convex sides, nearly 1903.] NATURAL SCIENCES OF PHILADELPHIA. 225 twice as long as broad; in profile the anterior and posterior slopes of the node are of nearly equal length, the former slightly concave, the latter distinctly convex, the apex somewhat flattened. Postpetiole hardly more than half as long as the petiole, but half again as broad, with prominent, rounded anterior angles and convex node. Head subopaque, longitudinally reticulate rugose, mandibles and frontal area smooth, clypeus with several longitudinal rugse, which are so delicate as to leave the surface shining. Thorax subopaque, irregu- larly reticulate rugose, especially on the pronotum and epinotum, the neck and pleurae still more delicately reticulate. Petiole and post- petiole subopaque, delicately reticulate rugose. Gaster smooth and shining. Hairs on the head and trunk whitish, rigid and obtuse, but hardly clavate, longest and most conspicuous on the aljdomen. Antenna and legs clothed with delicate, appressed, whitish hairs. Color yellow, whole head, except the mandibles, dark -brown. There is a large brown blotch on the pronotum and one on the epinotum. Nodes of petiole and postpetiole, trochanters, swollen portions of the femora, club of antennae and the dorsal surface of the gaster distinctly infuscated. Type locality: Boulder, Colo. Described from a dozen specimens received from Rev. P. J. Schmitt, O.S.B. This variety appears to be identical with the form mentioned by Emery from Hill City, S. Dak. (Pergande). It differs from specimens of the typical European miiscorum, sent me by Prof. Forel from the Schluderbach, Switzerland, in the following points: Hairs on the trunk distinctly shorter and at least on the head, thorax and pedicel distinctly thicker. Dorsal portions of thorax and pedicel and the thickened portions of the femora infuscated. These characters are constant in the twelve Colorado specimens. 3. Lepto thorax acervorum Mayr, subs, canadensis Provancher. L. canadensis Provancher. Addit. Faun. Canada, Hynienopt., 1SS7, p 245. §$c?. L. acervorum var. canadensis Er. Andre, Rev. d'EntomoL, A I, 1887, p. 295. ^ . L. acervorum var. canadensis Dalla Torre, Catal. Hvmenopt., VII, 1893, p. 123. L. canadensis Emerv, Zool. Jahrb. Abth. f. Svst., VIII. 1894, pp. 317, 318, 319. Worker (PI. XII, fig. 4). — Length 2.75-3.5 mm. Minute ocelli occasionally present. Cl3^peus with a distinct longi- tudinal impression in the middle, its anterior border rather convex and 15 226 PROCEEDINGS OF THE ACADEMY OF [Feb., rounded. Anteniije 1 1-jointed ; scape reaching midway between the eye and the posterior angle of the head; first funicular joint not longer than joints 2-3 together; terminal joint distinctly longer than the two pre- ceding joints of the club. Thorax long, subcylindrical, somewhat nar- rowed behind, dorsally flattened, its anterior angles rounded; pro- mesonotal and mesoepinotal sutures both distinct, a slight but distinct constriction at the latter. Epinotal spines rather short and blunt, compressed, in many specimens tooth-like and hardly longer than broad at the base, directed backward in a line with the dorsal surface of the epinotum, in no case longer than their distance apart at the base. Petiole from above oblong, not more than H times as long as broad, the node in profile with evenly concave anterior and somewhat de- pressed but convex posterior slope. Postpetiole trapezoidal, distinctly broader than long, its anterior angles prominent, anterior border distinctly broader than the posterior, dorsal surface hemispherical in profile. Mandibles shining, coarsely punctate. iMedian impressed surface of clypeus smooth and shining, lateral surfaces longitudinally rugose. Head opaque, longitudinally reticulate rugose, the rugse being most distinctly longitudinal on the front, vertex and cheeks. Thorax opaque, covered wdth reticulate rugae w^hich have a decidedly longi- tudinal trend on the dorsal and lateral surfaces of the pronotum and on the dorsal surface of the epinotum and the mesopleurce. On the mesonotum the reticulation resolves itself into an area of even, close-set foveolse. Pedicel opaque, the petiole sculptured like the mesonotum, the postpetiole somewhat smoother and almost punctate. Gaster smooth and shining. Hairs whitish; those on the head, thorax and pedicel short, rigid and subclavate, on the gaster longer and thinner; on the antennse and legs dehcate, appressed. Color dark-brown, almost black; small joints of the funiculus, the metatarsi, bases of femora, trochanters, ventral surface of pedicel, and in many specimens also the sutural regions of the thorax, yellow or pale-brown. In some specimens (immature?) nearly the whole thorax and the whole of the tibiae are yellow. Female (dealated). — Length 3.75-4 mm. Apart from the distinctively sexual characters, the female differs from the worker in the following characters : The reticulate rugosity of the head is more decidedly longitudinal, the rugae running back regu- larly to the occiput without deviation at the ocellar region. Meso- notum, scutellum and pleurte traversed by distinct longitudinal rugae. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 227 the first somewhat shining in the middle near its anterior border and in the regions of the parapsidal furrows. On the epinotum, about the bases of the spines, the rugse become very coarse and reticulate. Epi- notal spines short and blunt, resembhng those of the worker in shape and direction. Petiolar node pointed, with rather flat anterior and posterior slopes, coarsely reticulate rugose. Sculpture of postpetiole less pronounced than that of the petiole, but coarser than the petiole of the worker. Pilosity short, like that of the worker, but the hairs on the head, thorax and pedicel are less clavate. Head, thorax, abdomen, femora and antennal club, black, remaining portions of the legs and the funicle, brown. Type locality : "Canada." Additional locahties: Elk county. Pa. (Bradley); Olympia, Wash. (Kincaid). This form should, I believe, be regarded as belonging to the same species as the European acervorum, as suggested by Andre. As Emery claimed, however, it deserves to rank as a subspecies, and not as a variety. The workers differ from the European specimens of acer- vorum in my collection (from Switzerland (Forel) and Scotland (Dughch)) in the shorter epinotal spines, the much deeper coloration of the thorax, pedicel and legs, the shorter and more clavate hairs on the trunk and the minute, appressed, instead of suberect hairs on the antennal scapes and legs. My specimens of the North American form average nearly as large as the European. This subspecies is certainly rare in the Eastern States, but seems to be very common in Washington, to judge from the number of different nests sent me from that state by Prof. Kincaid. This is probably sig- nificant in connection with the palearctic distribution of acervorum. The habits of the American subspecies are unknowTi. They prob- ably resemble those of the European form, which lives in small colo- nies under bark, in moss, etc. 3a. Var. yankee Emery. L. canadensis Prov. var. yankee, Zool. Jahrb., VIII, '94, p. 319. ? 9 . The worker (fig. 5) differs from the worker of canadensis typ. in lighter coloration and in having somewhat longer epinotal spines. Head dark-brown, gaster somewhat paler; mouth, thorax, pedicel and legs reddish; antennal club, thoracic dorsum and femora usually infuscated. Sculpture finer and less rugose than in canadensis. In the female the thorax is dark-brown, the sculpture more pronounced. Type localities: South Dakota, Utah, Colorado. 228 PROCEEDINGS OF THE ACADEMY OF [Feb., Several workers sent me by Rev. P. J. Schmitt, O.S.B., from Boulder, Colo., agree very well with Emery's description. 3b. Var. convivialis var. nov. Length of worker 2-2.5 mm.; of female 3 mm. Differs from the typical canadensis and the preceding variety, in its small size and very deep coloration. Head, thorax, abdomen, femora and tibiae black, neck, ventral portions of pedicel, funiculus, trochanters, knees and tarsi red or yellow. Epinotal spines short and blunt. Sculpturing of body as rough as that of the typical canadensis. Color of the female deeper than that of the worker, the shining region of the mesonotum is more extensive than in the female of canadensis, and there is a large sliining area devoid of sculpture in the middle of the scutellum. Type locality : Milwaukee, Wis. Additional localities: Colebrook, Conn.; top of Las Vegas Range (11,000 feet), N. M. (T. D. A. Cockerell) ; Beulah, N. M. (F. W. P. Cock- erell). This variety seems to have a pronounced tendency to symbiosis with other species of Myrmicidae. The Milwaukee specimens were found living in the bark of a stump in xenobiosis with Cremastogaster lineolata Say. Those from the top of the Las Vegas Range were taken by Prof. Cockerell in a nest of Myrmica brevinodis. The Connecticut specimens appeared to be living in plesiobiosis with For- 7nica rufa subsp. difjicilis Emery. 3c. Var. Kincaidi Pergande. L. Yankee Emery var. Kincaidi Pergande, Proceed. Wash. Acad. Sci.,Vol. II, December 20, 1900, pp. 520, 521. ^ $ . "Female. — Length about 4 mm. "Head and thorax black, the abdomen dark-brown, with the pos- terior edge of the segments brownish-yellow; antennae, mandibles and legs yellowish-red, the neck and paler parts of the nodes of a darker red; the fiagellum grows gradually darker toward the end, with the last joint black; femora dark-brown, their base and apex yellowish-red; teeth of mandibles black. Head finely striated, the striae most distinct in front of the eyes and between the frontal carinae; the posterior half of the head is finely and rather densely rugose or reticulate, the clypeus is almost smooth and the mandibles striato-punctate ; pro- notum and mesonotum quite coarsely rugose, the metanotum, scu- tellum and upper surface of nodes finely, though rather indistinctly, striated; declivity of the metathorax transversely striated. Abdomen smooth. Erect hairs short, truncate and pale-yellowish, those of the nodes and abdomen longest; there are also a few much finer, erect . hairs on the femora. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 229 "The female resembles somewhat that of L. yankee, which, however, is somewhat smaller, the last antennal joint and metanotal spines shorter, the hairs of the abdomen much finer and the erect hairs of the femora wanting. "Worker. — Length about 3 mm. "Head and teeth of mandibles black, the abdomen dark-brown; an- tennae, mandibles, thorax, legs and nodes reddish-yellow; coloration of the last three or four joints of the antennae and the femora as in the female, the upper surface of the thorax and nodes more or less decidedly reddish-brown, Striation of the head more distinct than in the female, and the space between the striae more or less distinctly reticulated, particularly so toward the sides. Pronotum and mesono- tum and the nodes finely rugose; sculpturing of the metanotum slightly coarser. Abdomen smooth; all the hairs similar to those of the female. "The worker is very similar in appearance to those of L. yankee, though somewhat larger, more robust, the sculpturing coarser and the hairs stouter." Type locality: Metlakahtla, Alaska (June). Cat. No. 5,278 U. S. National Museimi. Described from one female and twelve workers. 4. Leptothorax Provancheri Emery. Myrmica tuberum Provancher, Natur. Canad., V, 12, 1881, p. 3592. Faune Entom. Canad., Hymenopt., 1883, p. 602. ? . Leptothorax Provancheri Emery, Zool. Jahrb. Abth. f. Syst., VIII, 1894, pp. 317 and 320. ? . Worker. — Length 2.75 mm. Body robust. Antennae U-jointed. Thorax stout, impressed at the mesoepinotal suture. Epinotal spines short, tooth-like, resem- bling those of L. canadensis var. yankee. Petiole with subparallel sides, node angulate above. Postpetiole about half again as^ broad as the petiole, transversely elliptical. Opaque, rugose-punctate; sculpturing like that of acervorum, the rugae on the head less numerous, forming wide meshes. Postpetiole very smooth and shining, with a few piligerous punctures, but other- wise impunctate. Gaster smooth and shining, with short longitudi- nal striae at its extreme base. Hairs rather long, clavate, both on the body and on the tibiae. Testaceous, crown of head infuscated. Type locality: "Canada" (Provancher). Redescribed by Emery from a single specimen, Provancher's orig- inal description being inadequate for the identification of the species. 230 PROCEEDINGS OF THE ACADEMY OF [Feb., 5. Leptothorax Emersoni Wheeler. L. Emersoni Wheeler, Am. Natural., XXXV, 1901, pp. 433-436. 5(j^$ . Worher (PL XII, fig. 6).— Length 2.5-3.3 mm. Head rather convex above, exchiding the mandibles not much longer than broad, sides rounded, posterior margin very faintly excised in the middle. Eyes rather large, convex, in the middle of the lateral surface of the head. Ocelli often present. Clypeus large, convex, broadly rounded in front. Mandibles 6-toothed. Antennae 1 1-jointed, scape reaching nearly to the posterior angle of the head; first funicular joint but little longer than the second and third joints together ; terminal joint not quite as long as joints 7-9 of the funiculus. Thorax rather long, rounded in front, narrowed behind, with distinct promesonotal and mesoepinotal sutures, the thorax distinctly con- stricted at the latter. Epinotal spines short, blunt, compressed, hardly longer than broad at their bases, directed somewhat upward and backward, their distance apart at the base greater than their length. Petiole from above oblong, 1^ tnnes as long as broad, sides slightly convex just in front of the middle ; in profile the node is pointed, with distinctly concave anterior and posterior slopes. In some speci- mens, however, the posterior slope is straight or even somewhat con- vex. Ventral portion compressed, produced forward as a blunt projec- tion. Postpetiole campanulate, wdth evenly convex dorsal surface; seen from above it is nearly twice as broad as the petiole, its anterior portion evenly rounded, not angulate. Gaster rather large, of the usual shape, without distinct anterior angles. Head opaque. Mandibles coarsely punctate. Clypeus smooth in the middle, with a few longitudinal ruga on either side. Head tra- versed by coarse longitudinal and occasionally anastomosing rugse; interrugal spaces coarsely and more or less confluently foveolate- punctate. Around the eyes the rugse become more reticulate, though they still have a distinctly longitudinal trend on the cheeks and lower surface of the head. Thorax opaque, its dorsal surface reseml^ling the head in sculpture, except that the rugse are more reticulate and without longitudinal trend. On the pleurse the rugse become indis- tinct and are replaced by even and closely aggregated foveolse. Petiole opaque, sculptured like the pleurse. Postpetiole smooth and shining above, delicately reticulate under a high magnification, especially on the sides. Gaster very glabrous and shining. Whole body, including the legs and antennae, abundantly beset with rather long, suberect, whitish hairs, which on the trunk are obtuse but by no means clavate. Hairs on the gaster longest and most regularly arranged. 1903.] NATURAL SCIENCES OF PHILADELrHIA. 231 Yellow or reddish-yellow, dorsal surface of head, terminal half of funiculus and the gaster, with the exception of the anterior, lateral and posterior borders of the segments, both on the dorsal and ventral surfaces, black or dark-brown. Female. — Length 2.5-3.5 mm. Like the worker. Thoracic dorsum dark-brown, rather shining; mesonotum and scutellum traversed by distinct, rather widely sep- arated, longitudinal rugoe; pronotum and epinotum coarsely reticu- late rugose, the rugae of the latter being continued up onto the dorsal and lateral surfaces of the stout, blunt spines. Pleurae subopaque, coarsely reticulate rugose, the meshes being occupied by aggregated shallow foveolae. Node of petiole somewhat more acute than that of the worker. Wings grayish hyaline, veins and stigma dirty yellowish. Male. — Length 2.5-3 mm. Mandibles very small, not distinctly dentate and far from meeting each other with their blades. Antennae 12-jointed; scape hardly as long as the three first joints of the funiculus together, first funicular joint very short, the others cylindrical, of uniform thickness but in- creasing graduallv in length toward the tip. Parapsidal and other thoracic sutures very distinct. Epinotum with two very short rugose projections in the place of the spines. Petiolar node low, rounded, its anterior slope slightly concave, its posterior slope shorter and con- vex. Postpetiole hemispherical. Gaster somewhat less flattened and narrower than in the worker. Mandibles striated. Clypeus even in the middle with a few delicate longitudinal rugae. Head above longitudinally reticulate rugose; rugae radiating backward and laterallj^ from the posterior ocelli as centers. Thorax rather smooth, indistinctly punctate. Pedicel and gaster glabrous. Hairs almost completely absent on the head and thorax, short and inconspicuous on the legs, long on the pedicel and gaster, but nowhere truncated at their tips. Brownish-yellow like the worker. Head, thoracic dorsum, pedicel and gaster, and the middle portions of the femora and tibiae somewhat darker. Wings as in the female. Type locality: Colebrook, Litchfield county. Conn. Males and females appearing in August. This species always lives in xenobiosis with a larger Myrmicine ant (Mijrmica hrevinodis Emer}^ in the hummocks of moss {Polytrichum commune), under stones, bits of wood, etc., in rather damp, grassy bogs. The Leptothorax occupy separate nests, which, however, com- 232 PROCEEDINGS OF THE ACADEMY OF [Feb , muiiicate by means of narrow passages with the galleries and chambers of the Myrinica. They obtain their food by licking the surfaces of the Myrmica and by regurgitation. All sorts of transitional forms occur between the Avorkers and queens (ergatoids and macroergates with from one to three ocelli). 6. Leptothorax Schaumi Roger. L. Schaumi Roger, Berl. Ent. Zeitschr., VII, 1863, p. 180, No. 70. ^ . L. Schaumi Mayr, Verb. Zool. bot. Ges. Wien, XXXVI, 1886, p. 451. O'. L. Schaumii Dalla Torre, Catalog Hvmenopt., VII, 1893, p. 127. L. Schaumi Emery, Zool. Jahrb. Abth. f. Syst., VIII, 1894, p. 320. Worker (PI. XII, fig. 7). — Length 2.5-2.75 mm, Mandibles 5-toothed. Clypeus convex without median impression; its anterior border rather straight. Antennae 11-jointed; scape reach- ing hardly to half way between the eye and the posterior angle of the head, funiculus with a distinctly 3-jointed club; first funicular joint almost as long as joints 2-5 together; joints 2-7 distinctly broader than long; terminal joint fully as long as the two preceding joints. Thorax rather short, flattened dorsally and laterally, broader in front than behind, with distinct and rather sharp humeral angles, and with a constriction at the mesoepinotal suture. Epinotal spines very short, dentiform, not longer than l3road at their bases. Petiole seen from above oblong, Ih times as long as broad, its sides parallel except at the peduncle which is narrower; in profile the anterior dorsal slope is concave and about the same length as the straight or somewhat con- vex posterior slope ; ventral surface with a distinct tooth directed for- ward. Postpetiole scarcely half again as broad as the petiole, distinctly broader than long, oblong, with distinct though rounded anterior angles. Its dorsal surface is evenly semicircular in profile. Gaster of the usual shape, with small but distinct anterior angles. Clypeus with sharp longitudinal rugse, two of which, near the middle, are more prominent than the others. ]\Iandibles with distinct longi- tudinal rugse. Head, clypeus and mandibles with a silky luster, the first traversed bj^ fine parallel rugse separated by rows of foveolate punctures, which are clearest in certain lights on the posterior lateral surfaces and cheeks. Thorax, petiole and postpetiole opaque, covered uniformly with foveolate punctures. Gaster smooth and shining. Hairs moderately numerous on the bod}^, white, erect, clavate; short on the head and thorax, much longer on the gaster and of inter- mediate length on the pedicel. Hairs on the antennae and legs minute, non-clavate, appressed. Yellowish-red, the edges of the mandibles black. Gaster in some specimens dark-ljrown throughout, in others yellow or with much of 1903.] NATURAL SCIENCES OF PHILADELPHIA. 233 the base of the first segment yellow. Antennae and legs yellow, club and sometimes also the scape of the former, infuscated. Male. — Length 3.2 mm. Mandibles dentate, touching each other with their blades. Antennae 12-jointed, scape about as long as the first three joints of the funiculus together; funiculus from the second joint to the end of uniform thick- ness, filiform; second joint a little shorter than the third, shorter, in fact, than anj^ of the succeeding joints. Instead of spines or teeth, the epinotum bears two indistinct elongate swellings. Radial cell of wings short and closed. Mandibles rather smooth and shining, with scattered punctures near their inner edges. Clypeus moderately shining and very deli- cately longitudinally rugose. Cheeks and region between antennal insertions and eyes sharply striated longitudinally; front with delicate longitudinal rugae; vertex finely reticulate punctate. Thorax rather smooth and shining, median and posterior portions of mesonotum finely longitudinally rugose and in part obliquely. Petiole, postpetiole and gaster smooth and shining. Pilosity sparse, tibiae without suberect hairs. Blackish-brown, pedicel and gaster darker. Mandibles, antennae except the brown scape, joints of legs, tarsi, and in part also the joints of the pedicel, yellow or reddish-yellow. Wings clear, hyaline. Type locality: "Pennsylvania" (Schaum). Additional localities: District of Columbia (Pergande); Beatty, Pa. (Schmitt); Westville, N. J. (Schmitt) ; Austin, Tex. In the locality last mentioned I have occasionally taken the workers of L. Schaumi running on the bark of large willows (Salix nigra). 7. Leptothorax fortinodis Mayr. L. fortinodis Mayr, Ver. Zool. bot. Ges. Wien, XXXVI, 1886, pp. 451, 452. ? ?. L. fortinodis Dalla Torre, Catalog. Hymenopt., "\ II, 1893, p. 124. L. fortinodis Emery, Zool. Jahrb. Abth. f. Syst., VIII, 1894, pp. 318, 321. Worker (PL XII, fig. S).— Length 2.5-3 mm. Head somewhat longer than broad, sides subparallel, occipital border nearly straight. Eyes moderately large and convex. Mandibles 5-toothed. Clypeus convex, its anterior border broad and rounded. Antennae 11-jointed; scape reaching to midway between the eye and the posterior angle of the head; funiculus terminating in a 3-jointed club; first funicular joint as long as joints 2-4 together; joints 2-5 broader than long, 6 and 7 as long as broad; terminal joint somewhat 234 PROCEEDINGS OF THE ACADEMY OF [Feb., longer than the two preceding joints of the club. Thorax moderately long, its dorsal and pleural surfaces flattened; humeral angles promi- nent and rather sharp; no constriction between the meso- and epinotum. Epinotal spines very short, not longer than broad at their bases, turned upward. Pedicel rather robust; petiole oblong Avhen seen from above, almost twice as long as broad ; sides of the nodal portion nearly parallel, peduncle narrower; in profile the anterior slope is distinctly concave, the posterior strongly convex, the apex of the node rounded; anterior ventral surface with a prominent tooth, directed forward and downward. Postpetiole but little broader than the petiole; nearly as long as broad, subglobose, strongly arched dorsally, its anterior about as broad as its posterior border, the anterior angles rounded, hardly distinct. Gaster of the usual configuration, with small but prominent basal angles. Legs robust. Mandibles and clypeus longitudinally rugose, the former finely and indistinctly, the latter more coarsely and distinctly. Head opaque, except the crown and occiput which are more or less shining or lus- trous, covered with foveolate punctures, in addition to which the front and crown are finely longitudinally rugose, the cheeks and sides reticulate-rugose, the lower surface more indistinctly reticulate. Thorax, petiole and postpetiole opaque, or slightly lustrous, densely and evenly foveolate-punctate. Gaster smooth and shining. Hairs on the body yellowish-white; erect and clavate on the head, thorax and abdomen; longest on the gaster; on the antennae and legs the hairs are minute, tapering and appressed. Head, thorax and pedicel deep reddish-brown, gaster almost or ciuite black; mandibles, antennae and legs red, club of antennae and often also the femora infuscated. Female (dealated). — Length 3.5—1 nun. Like the worker in coloration, except that the anterior and lateral portion of the mesonotum, the epinotum and dorsal surfaces of the petiolar and postpetiolar nodes may be distinctly infuscated. Whole upper surface of head longitudinally rugose, the rugae diverging to the corners of the head in the postocellar region. Mesonotum and scutellum traversed by numerous parallel longitudinal rugae, which are finer than those on the head ; pronotum finely reticulate, median por- tion of scutellum smooth and shining; pleurae reticulate-punctate, rather rough, the rugae somewhat longitudinal in direction. Epinotal spines very small and tooth-like, rather blunt ; the region between and below them densely foveolate-punccate. Petiole with less convex 1903.] NATURAL SCIENCES OF PHILADELPHIA. 235 anterior slope to the node; postpetiole relatively shorter and broader than in the worker. Type locality: Maryland. Additional locality: Austin, Tex. The specimens from Austin have the head, thorax and pedicel decidedly paler in color than a type specimen received from Dr. Mayr and three topotypes given me by Mr. Pergande ; the petiole is relatively smaller and there is a clavate hair at the base of each epinotal spine as in Schaumi. This hair is lacking in my t^'pes of fortinodis, possibly because these are much rubbed. I have found only a single colony of this form at Austin. This had taken up its abode in an abandoned gall of Holcaspis cinerosus Basset on the live-oak {Quercus virginiana). It contained 143 workers, a single dealated queen and 35 larvae in different stages. The latter were white and not greenish like the larvae of L. obturator which in- habits the same kind of galls. The entrance to the fortinodis nest was a small round hole with much worn edges, evidently the modified exit of some parasite on the Holcaspis. .'A. 7a. Var. melanoticus var. nov. Worker. — Length 2-2.5 mm. This form is decidedly smaller and much deeper in color than the typical fortinodis. Whole body black, petiole and postpetiole brown- ish behind and below; mandibles, joints of legs, tarsi and funiculus brown. Epinotal spines very short and blunt in some specimens, in others acute and longer, recalling the conditions described by Mayr for a small variety of fortinodis from the District of Columbia. Female (dealated). — Length 2.9-3 mm. Decidedly smaller than the typical form, head and thorax as well as the pedicel and gaster black ; the ventral surface of the petiole and post- petiole reddish. Legs dark-brown or black ; coxae, knees, tarsi and basal portion of funiculus somewhat paler. Type locality: Rockford, 111. The workers of this apparently somewhat depauperate form are sometimes seen running on the bark of large prostrate oak logs in the woods. They inhabit small flat chambers only ^ to f inch in diameter excavated in the thick corky bark. Each nest contains a single dealated queen and a comparatively small number of workers (about 25 to 40). n. Var. gilvus var. nov. Worker. — Length 2.25 mm. Differs from the typical fortinodis in the color, which is throughout 236 PROCEEDINGS OF THE ACADEMY OF [Feb., a clear yellow, and in the sculpturing of the head, which is smooth and shining except on the sides, where it is more opaque and reticulate. The vertex is traversed by a few rather widely separated rugse. Female (dealated). — Length 4 mm. Differs like the worker in coloration. The whole body is yellow, except the wing insertions, which are black. Type locality : Austin, Tex. This variety is based on a single dealated queen which was found accompanied by seven workers and a few larvae in a small Holcaspis cinerosus gall on a hve-oak tree. These evidently constituted an in- cipient colony, remarkable because the queen and two workers were pure yellow, while the five remaining workers were dark-brown with black gasters, like the workers of the typical fortinodis. All the workers, however, had the peculiar smoothness of the head and were undoubt- edly the offspring of the same mother. I believe this colony mvist present a case of hybridism, a female of the new variety gilva having been fertilized by a male of the typical fortinodis. It is difficult to explain the peculiar dichromatism of the workers in this little colony in any other way, since the yellow workers were not callows but per- fectly mature, and the queen differed so decidedly in color from the majority of her offspring. Unfortunately the colony was killed by dropping the gall into alcohol before the peculiarities of the workers were noticed. The Austin specimens of fortinodis, together with those representing the varieties melanoticus and gilvus, all have the petiole much smaller than in Mayr's type and suggest transitions to Schaumi. Particularly is this the case with var. gilvus, which is based on the female. The female of Schaumi and the males of both species being unknown, I am unable to dehmit the two species accurately. 8. Leptothorax longispinosus Roger. L. longispinosus Roger, Berl. Ent. Zeitschr., VII, 1863, p. ISO, No. 69. ^ . L. longispinosus ]Mayr, Verb. Zool. bot. Ges. Wien, XXXVI, 1886, p. 451. 5 • L. longispinosus Dalla Torre, Catalog. Hymenopt., VII, 1893, p. 125. L. longispinosus Emery, Zool. Jahrb. Abth. f . Syst., VIII. 1894. p. 321. ^ $ . Worker (PI. XII, fig. 9).— Length 2.2.5-2.5 mm. Head exclusive of the mandibles not much longer than broad, posterior angles considerably rounded. Mandibles 5-toothed. Eyes of moderate size, rather flattened, Clypeus convex, not impressed in the middle, its anterior border rounded. Antennse 11-jointed, scape reacliing the posterior angle of the head, club 3-jointed; first funicular joint as long as joints 2-4 together; joints 1-3 broader than, long; joints 4-6 as long as broad; terminal joint as long as the two pre- 1903.] NATURAL SCIENCES OF PHILADELPHIA. 237 ceding joints together. Thorax rather short, anterior angles rounded but distinct, dorsal surface and pleurae somewhat flattened; meso- epinotal suture distinct but without a constriction. Epinotal spines very long and stout, directed backward, rather suddenly tapering at their tips which are curved slightly inward and downward. Petiole from above suboblong, twice as long as broad, sides nearly parallel, posterior border a little broader than the anterior ; in profile the ante- rior slope is distinctly and evenly concave, the posterior convex; the anterior ventral surface has a distinct but rather blunt tooth; summit of node blunt. Postpetiole hardl}^ half again as broad as the petiole, as long as broad, with rounded but distinct anterior angles, convex dorsally. Gaster of the usual shape, with distinct anterior angles. Mandibles coarsely longitudinally rugose, hardly shining. Clypeus somewhat shining, traversed even in the middle by several clean-cut longitudinal rugse. Head shining, especially on the posterior and postero-lateral portions; anteriorly with clean-cut longitudinal rugse, which are coarsely reticulate and further apart on the cheeks, more delicate on the crown and occiput. Thorax opaque, except the meso- notum, which is somewhat shining. Neck coarsely and evenly punc- tate; remaining surface of thorax covered with coarse, irregularly longitudinal rugse w^hich extend up on the epinotal spines; interrugal spaces with shallow foveolate punctures, forming a secondary reticu- lation. Petiole and postpetiole opaque, coarsely rugose and punctate. Gaster very smooth and shining. Hairs silvery- wliite, those on the head, thorax and abdomen very regularly arranged, clavate, erect; on the antennse and legs minute, non- clavate and appressed. Head and gaster black; thorax and pedicel dark-brown; antennse and legs yellow ; scape and club of the former, coxse, femora and some- times also the tibise of the latter, infuscated. Mandibles dark-brown, their distal haK yellow. Female. — Length 3.5-4 mm. Head opaque, densely and rather finely longitudinally rugose. Eyes moderate; ocelli rather small. Thorax opaque; pronotum coarsely longitudinally rugose ; mesonotum traversed by numerous very regular, parallel rugse. ScuteUum somewhat shining, covered with much more delicate rugse than those of the mesonotiun and more reticulate and less longitudinal in direction. Pleurse, epinotum and epinotal spines covered with coarse reticulate rugse, which have a decidedly longitudi- nal trend. Epinotal spines shorter, stouter and less curved than those 238 PROCEEDINGS OF THE ACADEMY OF [Feb., of the worker. Petiole and postpetiole opaque and more roughly sculp- tured than those of the worker. Wings milky-white, the veins and stigma very pale. Pilosity and color of body, legs and antennse like the worker, except that the thorax is darker and often quite black, especially on the dorsal surface. Male. — Length 2-2.5 mm. Head, exclusive of the mandibles, about as broad as long. Eyes very prominent ; ocelli reniform. Mandibles overlapping, small, acute, dentate. Antenna 12-jointed; scape as long as joints 1-4 of the funiculus, the funiculus with a 4-jointed club; first funicular joint swollen, somewhat longer than joints 2-3 together; joints 3-7 cylin- drical, about twice as long as broad, joints of club fusiform gradually increasing in length distally. Thorax with strongly marked parapsidal and Mayrian furrows. Epinotum evenly rounded, with two small prominences in the place of the large spines of the worker and queen. Petiole larger and postpetiole more slender than in the worker and both with much lower nodes, the former somewhat pedunculate, the latter subquadrate from above, with rounded angles, as long as broad and hardl}^ half again as broad as the petiole. Gaster of the usual shape. Legs rather long and slender. Clypeus shining, with a few clean-cut, longitudinal rugse. Head subopaque, indistinctly rugose and punctate except the cheeks, where the rug£e are pronounced and reticulate. Thorax smooth; pleurse, mesonotum and scutellum shining, their surfaces indistinctly and irregularly punctate at the sutures. Epinotum opaque, very finely rugose. Petiole and postpetiole opaque, finely rugose; the upper sur- faces of the nodes, especially of the postpetiole, smooth and almost shining. Gaster subopaque. Hairs on the body few and very slender, whitish; longest on the gaster; those on the legs and antennae minute and appressed. Black; mandibles, antennae, legs and genitalia white. Bases of mandibles, scape, antennal club, coxse, femora, tibise and last tarsal joint of each foot, distinctly infuscated. Wings milky-white with very pale veins and stigma. Type locality : "America." Additional localities: Virginia (Mayr); District of Columbia (Per- gande) ; New York (Schmelter) ; Colebrook, Litchfield county, Conn. This species is evidently allied to L. curvispinosus, but is readily distinguished by its dark coloration, strong epinotal spines, shining head, etc. The specimens from which the above description was drawn may be considerably darker than Rogers' types. In most of J903.] NATURAL SCIENCES OF PHILADELPHIA. 239 my material the thorax of the worker is black, and the head rather smooth so as to resemble the forms described by Emery from New York and by Mayr from Virginia. L. longispinosus appears to he confined to the Eastern United States. At any rate I have not yet been able to find it in the ]\Iiddle West or among my material from the Western States. At Colebrook, Conn., the workers of this species are often seen running over the leaves or bushes in rather damp, shady places. The nests, containing the winged females and males in August, were found in clefts of granite boulders and in worm-eaten hickory nuts on the ground under the trees in the woods. Some of the colonies were quite populous for Lcptothorax colo- nies, others very small. 9. Leptothorax curvispinosus Mayr. L. curvispinosus Mayr, Sitz. B. k. Akad. Wiss. Wien, LIII, 1866, p. 508. ^ . Stenamma gallarum Patton, Am. Natural., 1879, p. 126. ^ 9 . L. curvispinosus Mayr, Verhand. zool. bot. Ges. Wien, XXXVI, 1886, pp. 451 and 453. '$ . L. curvispinosus Dalla Torre, Catalog. Hymenopt., VII, 1893, p. 124. L. curvispinosus Emery, Zool. Jahrb. Abth. f. Syst., VIII, 1894, pp. 317 and 320. Worker (PI. XII, fig. 10).— Length 2-2.5 mm. Mandibles 5-toothed. Clypeus moderately convex, with broadly rounded anterior border, without median impression. Antennae 11-jointed, scape reaching to midway between the eye and the posterior corner of the head; funiculus terminating in a distinct 3-jointed club; first funicular joint nearly as long as the three succeeding joints together; joints 2-7 about as long as wide; terminal joint little longer than the two preceding joints of the club taken together. Thorax but little broader in front above than below and behind; humeri shghtly angular; dorsum convex, without promesonotal and meso- epinotal sutures and without a constriction at the latter region. Epi- notal spines long and slender, tapering rather rapidly at their tips; directed backward and slightly upward, their tips incurved and slightly converging. Petiole from above more than twice as long as broad, distinctly narrower at the anterior peduncular end than behind; node in profile rather blunt, with longer and slightly concave anterior slope and convex posterior slope; lower surface laterally compressed, with a small but distinct downwardly directed tooth near the anterior end. Postpetiole globose, about half again as broad as the petiole, almost circular when seen from above. Gaster short, elliptical, with small but distinct basal angles. Mandibles shining, with indistinct longitudinal strise. Clypeus longitudinally rugose, even in the middle. Head opaque, covered 240 PROCEEDINGS OF THE ACADEMY OF [Feb., with fine longitudinal rugne separated b}^ densely foveolate punctures. Thorax opaque, coarsely and irregularly longitudinally rugose, except just back of the neck where there are a few transverse rug£e. The rugse on different parts of the thorax are so fine that their trend is hardly discernible except under a high magnification. Lower pleurae foveolate-punctate. Petiole and postpetiole somewhat less roughly sculptured than the head and thorax; both densely foveolate-punctate; punctures on the petiole somewhat coarser than on the postpetiole, which therefore often appears smoother. Gaster and legs smooth and shining. Hairs whitish, those on the head, thorax and pedicel shorter and more clavate than those on the gaster. Hairs on the legs and antennae short, non-clavate and appressed. Yellow; head, thorax and gaster tinged with l^rown; mandibles, legs, antennse and venter pale, sometimes whitish; edges of mandibles and a large triangular spot on either side of the first gastric segment, black or dark-brown. In some specimens the femora are slightly infuscated. Female (dealated). — Length 2.75-3.3 mm. Longitudinal rugse of the head more prominent than in the worker. Pronotum coarsely reticulate-rugose. Mesonotum shining, especially in front and in the parapsidal regions, longitudinally rugose, as are also the paraptera and scutellum. Epinotum with coarse, transverse rugse, especially below the spines, which are shorter and stouter than in the worker. Pleurae and sterna coarsely longitudinally rugose. Sculpturing of the petiole and postpetiole like that of the worker but more pronounced, so that these segments are quite opaque. Upper surface of head, scutellum, posterior portion of epinotum, wing-inser- tions, lower pleurae, posterior portions of petiole and postpetiole, a broad band across the first gastric segment and all except the borders of the posterior gastric segments, dark-brown or black. Type locality : ? District of Columbia. Additional localities: Virginia (Mayr) ; Beatty, Pa. (Schmitt) : Bel- mont, N. C. (Schmitt) ; Covington, Ky. (Schmitt) ; New York (Emery) ; New Jersey (Emery). This species in its typical form appears to be confined to the Eastern ITnited States. Patton found small colonies of it nesting in the hollow galls on the golden-rod (Soliclago). Rev. P. J. Schmitt, O.S.B., who has frequently taken the species in Pennsylvania, sends me the fol- lowing note on its habits: ''In one locality at least where curvispi- nosus was abundant the colonies were in saplings of ash, the tops of 1903.] NATURAL SCIENCES OF PHILADELPHIA. 241 which had been eaten off b}' cattle and hollowed out, perhaps by some larger insect than Leptothorax. At all events, when I visited these colonies in autumn (I knew of their existence by watching foraging workers going in and out of the nests) every colony had been dislodged and dispossessed of its premises by a species of wasp which was busily bringing in paralyzed spiders. The L. curvispinosus had then retired to hollows in stumps or logs or dead branches lying on the ground." This species is also of interest because it is enslaved by Tomognathus americanus Emery, in the nests of which it has l^een found b}^ Per- gande. 9a. Subsp. ambiguus Emery (I'l. XII, fig. U). L. curvispinosus Mavr subsp. ambiguus Emory, Zool. Jahrb. Abth. f. Syst., VIII, 1894, p. 320^. Differs from the t3'pical curvispinosus in the somewhat coarser sculp- turing and the decidedly shorter and nearly straight epinotal spines. Type locahties: Hillbity, S. Dak. (Pergande) ; Cleveland, O. (Was- mann); New York (Schmelter). A number of specimens collected at Colebrook, Conn., have the same sculpturing as the typical curvispinosus but decidedly shorter epinotal spines. These were found running on the surfaces of leaves in the shade of very damp woods. I failed to discover the nests. 9b. Subsp. rugatulus Emery (PI. XII, fig. 12). L. rugatulus Emery, Zool. Jahrb. Abth. f. Syst., YIII, 1894, p. 321. Owing to the existence of the next subspecies (anncctcns) , I feel justi- fied in regarding Emery's L. rugatulus merely as a subspecies of curvispinosus. The type specimens were from South Dakota (Per- gande) and Colorado (Pergande). Specimens from Seattle, Wash. (Kincaid), in my collection agree very closely with Emery's description. They differ from the typical curvispinosus in the following characters : The rugse of the head and thorax are decidedly coarser, and longitudi- nal on the latter. Epinotal spines much shorter and hardly curved. Tooth on the anterior ventral surface of the petiole distinctly larger, blunter and directed downward and forward. Postpetiole broader than long, oblong when seen from above, with rather distinct anterior angles. In profile the upper surface of the postpetiole is almost angu- lar, its ventral surface very short. Upper surface of head and gaster, with the exception of the posterior edges of the segments of thejatter, dark-brown. Femora more or less infuscated in some specimens. 9c. Var. Cockerelli var. nov. Worker. — Length 2-2.5 mm. Differs from the typical rugatulus in having the head, thorax~and 16 242 PROCEEDINGS OF THE ACADEMY OF [Fcb,. patiole less opaque, owing to the rugiB being further apart and the smooth interrugal spaces more prominent. Epinotal spines shorter and more acute. Postpetiole somewhat longer, being intermediate in shape between that of the typical curvispinosus and rugatulus, but distinctly wider in front than behind, and with rounded but percep- tible anterior angles. The coloration is also intermediate between the forms just mentioned. Upper surface of head pale-brown and more of the cheeks, sides and front of head yellow than in rugatulus, whereas the infuscation of the gaster is limited to the posterior dorsal half or two-thirds of the first segment, leaving the remaining segments yellow. Female (dealated). — Length 3-3.5 mm. Whole body, with the exception of the legs and antennae, yello\^^sh- brown; head and gaster, with the exception of the base of its first segment, darker. Antennae and legs more yellowish. Pedicel, espe- cially the petiole, very rough, and surmounted by a more acute node than in the worker. A fine living colony of this species, comprising more than a hundred workers and eight females, was sent me by Prof. T. D. A. Cockerell from Las Vegas Hot Springs, N. ^I. Fragments of bark accompany- ing the ants showed that the nest was found in a tree trunk. 9d. Subsp. annectens subsp. nov. (PI. XII, fig. 13). Worker. — Length 2-2.5 mm. This form has the epinotal spines long and thin, and shaped like those of the typical eurvispiiiosus. Head very coarsely longitudinally rugose, subopaque. Pronotum evenly and coarsely foveolate-punctate, meso- and epinotum opaque, coarsely reticulate rugose, the rugae without a longitudinal trend. Petiole and postpetiole opaque, shaped like those of rugatulus, the former with a prominent ventral tooth, directed down- ward and forward. Upper surface of head and whole dorsal surface of gaster, except a large, transversely elliptical spot on the anterior portion of the first segment and the extreme posterior edge of this and the remaining segments, dark-ljrown or black. Remainder of body brownish-yellow. Type locality : Boulder, Colo. Described from four specimens collected by Rev. P. J. Schmitt, O.S.B. This form is clearly intermediate in structure and coloration be- tween the typical curvispinosus and the subspecies rugatulus. 10. Leptothorax Schmittii sp- uov. Worker (PL XH, fig. 14).— Length 2-2.25 mm. Head conspicuously narrow, with parallel sides, decidedly longer 1903.] NATURAL SCIENCES OF PHILADELPHIA. 243 than broad, with straight posterior border. Mandibles 5-toothed. Clypeus moderately convex, its anterior border with a small but dis- tinct excision in the center and a distinct median carina extending nearly its full length. Antennae 12-jointed; scape reaching the pos- terior corner of the head; first funicular joint as long as joints 2-4 to- gether; second joint as long as broad ; joints 3-8 nearly as long as broad; club distinctly 3-jointed, first and second joints subequal, together a lit- tle shorter than the terminal joint. Thorax rather long and narrow, widest in front where the humeral angles are sharp and prominent. In profile the pronotum rises very abruptly from the neck, so that a trans" verse ridge is formed which gives the thorax a square-shouldere appearance; dorsum flatly and evenly rounded, without mesoepinotal constriction. Epinotal spines well-developed, longer than broad at their bases, tapering and pointed, directed distinctly backward though slightly upward and outward ; in profile the ventral outline of the spines is distinctly concave, the dorsal convex. They are about as long as their distance apart at the base. Petiole about 1^ times as long as broad, distinctly broader behind than in front when seen from above; in profile the height of the node is fully equal to the length of the whole joint; its anterior surface is steep and somewhat concave, the top of the node abruptly truncated, the posterior slope so steep that it is even inclined forward below and forms somewhat less than a right angle with the extreme posterior dorsal surface of the petiole; ventral tooth well developed, directed downward. Postpetiole hardly twice as broad as the petiole, distinctly broader than long ; its anterior wider than its posterior border, its anterior angles rather prominent. Gaster of the usual shape. Mandibles not distinctly striated; shining, with a few coarse punc- tures. Clypeus shining, longitudinally rugose on the sides. Head very smooth and shining, covered with rather coarse but sparse piligerous punctures ; sides of frontal area, cheeks and subocular region subopaque and delicately longitudinally rugose. Pro- and mesonotum very smooth and shining, with a few piligerous punctures passing over onto the pleurae into delicate longitudinal rugae, which become much coarser and distinctly reticulate on the sides and whole upper surface of the epinotum. Petiole and postpetiole opaque, reticulate and punctate-rugose. Gaster very smooth, shining. Hairs white, only moderately abundant; clavate and erect on the thorax and crown of head, somewhat longer and more reclinate on the pedicel and gaster; the hairs on the sides of the head, antennae and legs non-clavate, appressed; those on the clypeus thin and projecting. 244 PROCEEDINGS OF THE ACADEMY OF [Feb., ^'ery dark-brown, almost black. ]\Iandibles, neck, funiculus and legs yellow; scape and club of antennae and the middle of the femora and tibise infuscated; edges of mandibles black. Type locality: Canon City, Colo. Described from four specimens collected by Rev. P. J. Schmitt, O.S.B., to whom I take pleasure in dedicating this very striking species. It is quite unlike any of the other described North American forms in the shape of the thorax and petiole and the smoothness of the head, pro- and mesonotum. 11. Leptothorax nitens Emery. L. nitens Emery, Zool. Jahrb. Abth. f. Syst., VIII, 1S94, pp. 318, 322, 323. Worker (PI. XII, fig. 15).— Length 2-2.25 mm. Mandibles 5-toothed; basal teeth very small. Clypeus moderately convex, impressed in the middle and with sinuately excised anterior border. Antennae 12-jointed ; scape reaching to | the distance between the eye and the posterior angle of the head ; funiculus terminating in a distinctly 3-jointed clul), the two basal joints of which are subequalin length, together decidedly shorter than the terminal joint ; first funicular joint as long as the three succeeding joints together, joints 2-7 of the funiculus slightly broader than long, the 8th about as long as broad. Thorax slender, somewhat broader in front than behind ; compressed laterally, pronotum rather prominent and square in front, humeri rounded ; in profile the dorsal surface is somewhat flattened and with- out mesoepinotal constriction. Epinotal spines very small, tooth- like, hardly as long as broad at their bases, directed upward, about as far apart as they are broad at their leases. Petiole about 1^ times as long as broad, gradually widened behind when seen from above; in profile the node is very high, its anterior slope steep and concave, its summit very short and rounded, the posterior slope abrupt, the ven- tral tooth is distinct and pointed forward and downward. Petiole nodiform, a little broader than long, half again as broad as the petiole, its anterior angles much rounded. Gaster of the usual shape. Mandibles smooth and shining, indistinctly striated and punctate. Clypeus smooth and shining, with a few longitudinal rugse on its antero- lateral surfaces. Head very smooth and shining, minutely and sparsely punctate; sides of front, antennal fovese and cheeks longitudinal rugu- lose. Thorax, petiole and postpetiole opaque, finely and regularly foveolate-reticulate. In some specimens more or less of the pro- and mesonotum is shining . Pleurse faintly striated longitudinally. Gaster very smooth and shining. Hairs moderately abundant, yellow; clavate on crown of head, thorax 1903.] NATURAL SCIENXES OF PHILADELPHIA. 245 and abdomen ; erect and shorter on head and thorax ; longer and slightly reclinate on the pedicel and gaster; hairs on sides of head, legs and an- tennse minute, non-clavate, appressed. Yellow, in some specimens crown of head, antennal club and dorsum of gaster slightly infuscated. Type locaUty: American Fork Caiion, Utah (Pergande). Additional locaUties: Pacific Grove, Cal. (Dr. H. Heath); Canon City, Colo. (Rev. P. J. Schmitt, O.S.B.). Emery described the species from a single specimen in which the thorax was shining. Examination of a number of California and Colorado specimens from the same nests shows this to be a common but by no means constant character; in a great many individuals the thorax is uniformly opaque throughout. One of the colonies sent me by Dr. Heath was found in the ground, hibernating in a Termite burrow. lla. Var. Heathii van nov. Worker. — Differs from the preceding in the coloration, which is constant in a whole colony sent me by Dr. Harold Heath from Pacific Grove, Cal. The body is brown, often rather dark, the legs and an- tennae brownish-white without the distinct yellow cast of the typical form. Found nesting in the ground under a stone. lib. Subsp. occidentalis subsp. nov. Worker. — This form combines the color characters of the type and the preceding variety. The ground color is yellow, the upper surface of the head, thorax and pedicel brown; the first gastric segment with a broad, brown dorsal band across its posterior half or two-thirds. Thorax and pedicel decidedly opaque. Antennal scape nearly reach- ing the posterior angle of the head, epinotal spines decidedly more robust than in the typical nitens and the var. Heathii. Type locahty: Friday Harbor, Wash. Described from six specimens received from Prof. Trevor Kincaid, 12. Leptothorax texanus sp. nov. Worker (PL XII, fig. 16).— Length 2.25-2.75 mm. Head longer than broad. Eyes rather large. Mandibles 5-toothed. Clypeus moderately convex, its anterior border somewhat rounded. Antenna 12- jointed; scape reaching nearly to the posterior angle of the head ; funiculus terminating in a very distinct 3-jointed club, the last joint of which is somewhat longer than the two preceding joints; first funicular joint nearly as long as joints 2-5; 246 PROCEEDINGS OF THE ACADEMY OF [Feb., second funicular joint about as long as broad, the remaining joints broader than long. Thorax rather short, its humeral angles rounded, dorsum evenly and slightly rounded, without mesoepinotal constriction. Epinotal spines moderate, distinctly shorter than the decUvous sur- face of the epinotum, scarcely as long as their distance apart at the base, rapidly tapering, acute, directed upward, outward and backward, their tips slightly deflected. Petiole from above fully Ih times as long as broad; its sides somewhat convex, so that its outline is sub- elliptical; in profile the anterior slope is abrupt and distinctly concave, the summit of the node flattened, and the posterior slope suddenly declivous; ventral tooth small, acute, directed downward. Post- petiole very large, fully twice as broad as the petiole; broader than long, its anterior and posterior angles rounded, so that it appears transversely elliptical from above; in profile it is very convex, espe- cially in front. Gaster elliptical, depressed, without distinct anterior angles. Mandibles coarsely striated. Clypeus traversed by several clean-cut longitudinal rugae, subsiding on the posterior portion which is smooth and shining. Head subopaque, coarsely reticulate-rugose; the rugae distinctly longitudinal only on the front and vertex ; interrugal spaces secondarily reticulate. Neck evenly reticulate; thoracic dorsum very coarsely and irregularly reticulate-rugose; the spaces between the rugae smooth and shining, because the secondary reticulation is indis- tinct or lacking. Pleurae somewhat more delicately and evenly rugose. Petiole and postpetiole coarsely reticulate-rugose, interrugal spaces filled with shallow foveolae or secondary reticulation. Gaster smooth and shining. Hairs white, rather numerous and prominent; on the trunk clavate; shorter and more erect on the head and thorax, longer and slightly reclinate on the pedicel and gaster; on the legs and antennae distinct, non-clavate, more or less appressed. Black or very dark-brown, especially on the head, thorax and petiole. Mandibles, excepting the teeth, funiculus, tips of epinotal spines, tarsi and articulations of legs, excepting the last tarsal joint, yellow; an- tennal scape brown. Female. — Length 3.75-4 mm. Head more decidedly and extensively longitudinally rugose than in the worker. Neck delicately, pronotum more coarsely reticulate- rugose; mesonotum and paraptera traversed by numerous longitudinal rugae which are more or less interrupted and have a tendency to anas- tomose. Scutellum with such rugae only in front and on the sides, 1903.] NATURAL SCIENCES OF PHILADELPHIA. 247 posteriorly it is nearly smooth and shining. Pleura^ and cpinotiim with rather coarse, distinctl)^ longitudinal rugae. Epinotal spines small, acute, straight. Petiole with pointed node and more abruptly decliv- ous posterior slope than in the worker, its dorsal surface scarcely flattened. Postpetiole very convex in front above. Sculpturing of petiole and postpetiole as in the worker. Color of head, thorax and pedicel reddish-brown, darker on the dorsal surface. Gaster black. Pilosity of body and legs and color of the latter much as in the worker. Wings whitish-hyaline; veins and stigma yellow. Male. — Length 2-2.5 mm. Head exclusive of the mandibles broader than long, cheeks short, posterior angles convex and rounded. Mandibles overlapping each other. Antennae 13-jointed; scape as long as the first four joints of the funiculus, club very distinct, 4-jointed, the first, second and third subequal, fully H times as long as broad, apical joint as long as the second and third together; first funicular joint somewhat swollen, fully Ih times as long as broad; joints 2-8 longer than broad. Epino- tum with two very small, indistinct protuberances in the place of the spines. Petiole slender, in profile concave below; anterior dorsal slope nearly straight, gradual, posterior slope shorter, abruptly de- clivous, rather concave. Postpetiole campanulate, about as long as broad; H times as broad as the petiole. Legs long and slender. Mandibles striated. Clypeus coarsely and irregularly reticulate- rugose, shining. Head subopaque, evenl}^ and rather delicately reticu- late-rugose. Neck finely reticulate; mesonotum shining, with a faint reproduction of the irregular rugosity of the worker; remainder of thorax and the pedicel delicately reticulate-punctate. Gaster smooth and shining. Hairs white, non-clavate, long and moderately abundant on the mandibles, pedicel and gaster; sparser elsewhere. Black; mandibles and genitalia yellow; antennal funiculus grayish- brown; wings like those of female; legs like those of the worker in color. Type locality: Milano, Millan County, Tex. All three sexual phases of this species were taken May 23, 1902, in nests consisting of a few small galleries, 3-4 inches long, excavated in .the sand in rather damp spots under post-oaks and cedars. 13. Leptothorax tricarinatus Emery. L. tricarinatus Emery, Zool. Jahrb. Abth. f. Syst., VIII, 1894, pp. 318, 321, 322. 5. Worker (PI. XII, fig. 17). — Length 2.25 mm. Clypeus produced in front in the middle, obtusely angulate, above 24S PROCEEDINGS OF THE ACADEMY OF [Feb., with three slender longitudinal carina?. Antenna? 12-jointed, first fiuiicular joint longer than the three succeeding joints together; remaining joints shorter than broad; two basal joints of club siibequal. Thorax not impressed at the mesoepinotal suture. Epinotal spines short, rather acute, ol^liquely erect. Petiole thickened behind, node subconical, obtuse. Postpetiole much larger than the petiole, sub- globose. Mandibles striated. Clypeus shining. Head, thorax and pe tiole subopaque. punctate and finel}' rugose, pronotum more shining in the middle. Hairs on the body clavate, on the legs and antennal scape non- clavate and sparse. Fuscous black; mandibles, articulations of legs and the tarsi reddish. Type locality: Hill City. S. Dak. (Pergande). Descriljed from a single specimen in the collection of Prof. C. Emery. 14. Leptothorax neomexicanus sp. nov. Worker (PI. XII, fig. 18).— Length 2.25-2.5 mm. Head longer than broad. Mandibles 5-toothed. Ch-peus mod- erately convex, broadly rounded in front, not impressed, Antennte slender, 12-jointed; scape extending to a distance equal to its own breadth beyond the posterior angle of the head ; funiculus terminating in a 3-jointed club, the two basal joints of which are subequal and together shorter than the terminal joint ; first funicular joint as long as the three succeeding joints together; joints 2-8 of the funiculus as long as broad. Thorax not much broader in front and above than behind and below; humeral angles rounded; dorsal surface flattened, without mesoepinotal suture or constriction. Epinotal spines short, robust, blunt, not longer than broad at their bases and nearly twice as far apart at their bases as long. They are directed obliquely upward, outward and backward. Petiole hardly 1^- times as long as broad; distinctly wider behind than in front, sides somewhat convex; in profile the node is nearly as high as the length of the petiole, its anterior slope steep, slightly concave; the summit somewhat truncated, passing abruptly into the angular posterior declivity ; ventral surface in front with a prominent tooth, directed downward and forward. Postpetiole nearly twice as broad as the petiole; nearly as long as broad, sub- globular, its anterior dorsal surface in profile abruptly convex, the posterior dorsal surface more flattened. Gaster of the usual shape with slight basal angles. Mandibles coarsely striated and punctate. Clypeus longitudinally rugose, especially on the sides, behintl without rugie, smooth and 1903.] NATURAL SCIENCES OF PHILADELPHIA. ^249 shinino-, somewhat foveolate-punctate. Head subopaque except along a broad median strip, extending from the frontal area to the occiput, and on the posterior angles. These regions are smooth and shining. Clieeks and sides of head rather delicately and longitudinally reticu- late-rugose. On the front and vertex there are also a few rather large indentations at widely separated intervals. Thorax nearly opaque, in front delicately and evenly reticulate-rugose ; on the epinotum and pleurse the rugte are coarser and have a distinctly' longitudinal trend. Petiole and postpetiole nearly opaque, finely reticulate-rugose. Gaster very smooth and shining. Hairs rather sparse, sil very- white ; those on the body subclavate, shorter and more erect on the head and thorax, longer and more reclinate on the pedicel and gaster. Hairs on the legs and antennie short, non-clavate, appressed. Black. In some specimens the pedicel and thorax are dark-brown. Scape and funiculus of antennae brown. Mandibles yellow, with black teeth. Legs yellow, middle portion of the femora and tibiae and terminal tarsal joint on each foot, black. Type locality: Manzanares, N. M. Described from five specimens taken by Miss Mary Cooper. The species is obviously closely related to L. tricarinatus Emery, and may prove to be merely a subspecies of this form. To judge from Emery's description, the head of tricarinatus is more opaque, the mesonotum shining, the first funicular joint larger than the three suc- ceeding joints, the remaining joints of the funiculus shorter than broad. Apparently, also, the postpetiole is considerably larger than in neo- mexicanus. 15. Leptothorax obturator sp. nov. Worker (PL XII, fig. 19).— Length 2.25-2.75 mm. Mandibles 5-toothed. Clypeus rather flat, its anterior margin broadly truncated in the middle. Antennae 12-jointed ; scape reaching posterior angle of head; first funicular joint as long as the three suc- ceeding joints; joints 3-S broader than long, joints 9-11 forming a club, the ninth distinctly narrower and shorter than the tenth, the terminal joint longer and considerably thicker than the two preceding joints. Thorax slender, somewhat broader in front than behind; rounded at the humeri, in profile convex in front and slightly concave behind on the dorsal surface, without mesoepinotal constriction. Epinotal spines small, rather acute, not longer than broad at their bases, nor further apart than long, directed upward. In front and 250 PROCEEDINGS OF THE ACADEMY OF [Feb., on the side of each epinotal spine there is a distinct longitudinal swell- ing or ridge on the epinotum. Petiole veiy slender, three times as long as broad, sides of the node parallel, the peduncle somewhat nar- rower when seen from above; in profile the lower surface is evenly concave, the anterior tooth minute; the dorsal surface with a low regularly rounded, knoll-like node. Postpetiole small, about H times as broad as the petiole, as broad as long, square when seen from above, with prominent anterior angles; in profile the low'er surface is flattened, the upper convex, especially in front. Gaster with distinct anterior angles. Sting well developed. Mandibles opaque, striated and with a few coarse punctures. Cly- peus subopaque, its whole surface longitudinalh^ rugose. Head sub- opaque in front and on the sides, shining behind, on the former regions densely and rather finely reticulate-rugose; the ruga? with a distinct longitudinal trend except on an opaque patch above each e^^e, where the rugge are evenly reticulate. The shining portion of the head is traversed by clean-cut longitudinal rugse much farther apart than on the front and crown and interspersed with a few coarse punctures; posterior angles of head delicately reticulate. Thorax subopaque; pronotal region more shining, foveolate-reticulate on the whole dorsal surface; pleurae more coarsely reticulate-rugose, with pronounced longitudinal trend in a few of the rugae, especially in the upper meso- and low^er metapleurse. Petiole and postpetiole subopaque, rather evenly foveolate-reticulate. Gaster smooth and shining. Hairs moderately abundant, snoAv-white; clavate on the crown of the head, thorax, pedicel and gaster; a little longer on the pedicel and gaster and somewhat more reclinate. Hairs on the legs and antennae sparse, non-clavate, appressed and inconspicuous. Black or very dark-brown. Mandibles, joints 3-9 of the antenna?, neck, ventral surface of petiole and postpetiole reddish-yellow'. Legs reddish-yellow, except the middle of the coxae, femora, tibiae and last tarsal joint, which are black. Female. — Length 3.5-3.75 mm. Clypeus and head more coarsely longitudinally rugose and more opaque than in the worker. Thorax subopaque; neck delicately reticulate-rugose ; pronotum transversely and irregularly rugose ; meso- notum, paraptera and scutellum traversed by dense, parallel, clean-cut, longitudinal ruga?. Epinotum delicately and somewhat concentrically reticulate-rugose ; armed with two inconspicuous swellings in the place of the spines; pleurae longitudinally rugose. Petiole and postpetiole like those of the worker, the node of the former relatively lower; post- 1903.] NATURAL SCIENCES OF PHILADELPHIA. 251 petiole hardly Ij times as broad as the petiole. The white hairs on the body are non-clavate, though those on the pedicel and gaster are somewhat thickened. Head, thorax and pedicel yellowish-red or dark-red, their upper surfaces more or less infuscated. Gaster black. Antennal scape black with basal half yellow; first joint of funiculus and club black, remaining joints yellow. Legs colored like those of the worker. Wings milky-white, veins yellow ; stigma brown, conspicuous. Male. — Length 1.5-2 mm. Head as long as broad, exclusive of the mandibles ; cheeks very short. Mandibles meeting with their tips. Clypeus convex, truncated in front. Antennse slender; 13-jointed; scape nearly as long as the first five joints of the fiagellum; first flagellar joint thickened, nearly as long as the three following joints together; joints 2-8 of the fiagellum cylindrical, as long as broad; the four terminal joints forming a club, of which the three basal joints are subequal in length but increase somewhat in thickness distally; terminal joint much larger, distinctly longer than the two preceding joints. Epinotum with two very in- conspicuous thickenings in the place of the spines. Petiole and post- petiole similar in shape to the corresponding segments of the worker, node of former very low. Clypeus shining, with a few reticulate rugse forming rather large meshes. Head, thorax and pedicel opaque, finely and evenly reticu- late-rugose. Mesonotum sparsely foveolate-punctate and traversed by a narrow, smooth, longitudinal stripe. Pleurse shining in part. Gaster smooth and shining. Hairs white, sparse, non-clavate, most abundant on the thoracic dorsum, pedicel and gaster. Black; pleural and pedicel more piceous. Mandibles, legs and antennse white; the mandibles w^ith brown edges, the antennse with scape, second joint and club blackened; legs with the coxse, middle of the femora and tibise and the last tarsal joint blackened. Type locality: Austin, Tex. Described from many specimens collected at different times from the abandoned Holcaspis cinerosus galls on the live-oaks {Q. vir- giniana). The young fertilized queen, on entering the gall to estab- lish her colony, gnaws n)inute fragments from the ligneous wall, mixes these wdth some secretion (saliva?) and completely plugs up the round opening through which the Holcaspis escaped and she herself has entered. Later when the first batch of tiny workers appear, they per- forate the center of the plug with a small opening like a pin-prick, and .just large enough for egress and ingress. This opening is too small for 252 PROCEEDINGS OF THE ACADEMY OF [Feb., the queen to pass, so that she remains imprisoned. With the growth of the colony the chamber formerly made and inhabited by the Hol- caspis larva is enlarged by the workers. The queen with the larvae prefers to inhabit the small central capsule in which the Holcaspis passed its pupal life. The larvae are of a peculiar greenish tint. The males and virgin females make their appearance in the colony during the last week of May. At no time are the colonies of L. obturator very large. They rarely comprise more than 36 or 40 workers. Members of different colonies, even from galls on the same branch, are extremely hostile to one another. Along the creek-bottoms near Austin, L. obturator is also occasionally found nesting in the twigs of the wafer- ash (Ptelea trijoliata) which have been hollowed out by tiny carpen- ter-bees {Ceratina nanula Ckll. and C. arizonensis Ckll.). The relatively large entrance made by the bees at the end of the twig is plugged up by the ants with agglutinated vegetable particles and then perforated with a minute opening in the center. 16. Leptothorax nevadensis sp. nov. Worker (PI. XII, fig. 20).— Length 2.5-3 mm. ^landibles 5-toothed. Clypeus depressed in the middle, its an- terior margin sinuately excised. Antennae 12-jointed, scape reaching to posterior corner of the head, first funicular joint as long as joints 2-4 together, joints 2-6 slightly broader than long, joints 7 and 8 as long as broad; two basal joints of club subequal, together shorter than the terminal joint. Thorax above in front of about the same width as below and behind; humeri much rounded, dorsum flattened in pro- file, without mesoepinotal constriction. Epinotal spines robust, pointed, decidedly longer than broad at their bases, and nearer together at their bases than long, directed upward, outward and backward. Petiole H times as long as broad, sides of node rounded, broader than the peduncle; seen from above the node is transverse, in profile it is narrow antero-posteriorly, its anterior slope gently ascending, concave, its posterior slope more abrupt, also concave, summit rounded; ventral surface of petiole with a prominent, compressed, downwardly directed tooth. Postpetiole in profile with a prominent, sometimes sHghtly angular node; the segment seen from above is transversely elliptical, about half again as broad as long, its anterior angles rounded. Gas- ter and legs of the usual conformation. Mandibles striate and punctate. Clypeus subopaque, its surface, especially at the sides, traversed by rather coarse longitudinal rugse. Head with a satiny luster, sparsely punctate and with dehcate longi- tudinal rugae, which become decidedly reticulate in the antennal 1903.] NATURAL SCIENCES OF PHILADELPHIA. 253 foveae. Thorax subopaque, its dorsal surface irregularly reticulate- rugose and foveolate, pleurae, petiole and postpetiole regularly foveo- late-punctate ; posterior epinotal declivity rather coarsely longitudi- nally rugose. Gaster smooth and shining. Hairs yellow, not very abundant; clavate on thorax, pedicel and crown of head ; short and erect on head and thorax, longer and some- what reclinate on the pedicel and gaster. Hairs on antennje, legs and sides of head less conspicuous and appressed, except on the antennae, where they are suberect. Rather dark reddish-brown, ventral portions of head, thorax and pedicel and the incisures of the gastric segments, yellowish. Legs and antennae yellowish, middle of femora and the antennal club dark- brown. Immature specimens have the thorax and pedicel more extensively yellow. Female (dealated). — Length 4.5 mm. Mandibles densely striated and somewhat punctate. Clypeus with coarse longitudinal rugae, one of which forms a distinct carina in the middle of the sclerite. Frontal area opaque. Head with coarse and very regular longitudinal rugae, but little diverging behind and but slightly reticulate. On the upper surface of the head there are also a number of shallow but distinct foveolae in the interrugal furrows. Pronotum and pleurae coarsely longitudinally rugose; mesonotum shining, sparsely foveolate and rather indistinctly longitudinally rugose, especially in front. Scutellum and paraptera like the meso- notum, the former with indications of rugae only at its anterior border. Whole epinotum subopaque, coarsely rugose, even over the entire surface of the robust, pointed spines, which are as long as they are broad at their bases. Declivous surface of epinotum regularly trans- versely rugose. Petiole and postpetiole opaque, reticulate and punc- tate-rugose, more coarsely on the sides than on the summits of the nodes ; petiolar node in profile more acute than in the worker. Gaster very glabrous. Head, thorax and pedicel rich reddish-brown; gaster decidedly darker. Legs and antennae yellow. In the latter all the joints of the funiculus are distinctly longer than broad, and the club, which is not infuscated, is indistinct. Wing-insertions black, Hairs on the body sparse, yellow, not clavate like those of the worker, but more or less tapering. Male.— Length. 2.5-3 mm. Mandibles dentate, overlapping with their blades. Clypeus convex, truncated in the middle in front. Antennae 13-jointed; scape slender, as long as joints 1-5 of the funiculus. First funicular joint fully twice 254 PROCEEDINGS OF THE ACADEMY OF [Feb., as long as broad, much stouter than the succeeding joints, except those of the 4-jointed chib; joints 3-7 longer than broad; three basal joints of club subequal, each not more than half as long as the terminal joint. Cheeks short. Thorax with very deep Mayrian and other sutures. Epinotum without indications of spines. Petiole and post- petiole longer and with lower nodes than in the worker. Gaster of the usual shape. Mandibles and clypeus somewhat shining, the latter with a few prominent and irregular longitudinal rugse. Head decidedly opaque, uniformly and densely punctate. Thorax shining, mesonotum, scu- tellum and pleurae with faint, parallel, longitudinal striae. Pronotum and epinotum more reticulate-rugose. Petiole and postpetiole smooth and shining on the summits of the nodes, elsewhere subopaque, finely reticulate-rugose. Gaster smooth and shining. Hairs covering the body sparse, whitish, non-clavate, longest and most conspicuous on the gaster, very small and appressed on the legs and antennae. Black, thorax and pedicel more piceous, especially on their lateral and ventral surfaces. The following parts are yellowish, or yellow suffused with piceous: mandibles, except their teeth, which are black, antennae, legs and genitalia. Wings whitish-hyaline, veins and stigma colorless. Type locality : King's Cafion, Ormsby county, Nev. Described from eight workers, one female and eight males, found by ]\Ir. C. F. Baker during August, 19n2, nesting in the ground under a stone. This species appears to be similar to L. tricarinatus Emery, but differs in the shape of the clypeus, which is impressed and sinuately excised and not produced, and in the shape of the epinotal spines, petiole, etc. 17. Leptothorax terrigena sp. nov. Worker (PL XII, fig. 21).— Length 1.5-1.75 mm. Head rather narrow, with parallel sides. Mandibles 5-toothed. Clypeus moderately convex, without median impression and with broadly rounded, non-sinuate anterior border. AntennjB 12-jointed, scape extending to f the distance between the eye and the posterior corner of the head; first funicular joint as long as the three succeeding joints together; joints 2-8 of funiculus decidedly broader than long, subequal ; three terminal joints forming a distinct club, of which the two basal joints are subequal in length but not in thickness, and together shorter than the terminal joint. Thorax slender, somewhat broader in front than behind, with rounded humeral angles and moderatelv 1903.] NATURAL SCIEN'CES OF PHILADELPHIA. 2d5 elevated anterior pronotal border. Pleurae compressed; dorsum flat- tened and without mesoepinotal constrictions. Epinotal spines small, conical, distinctly longer and further apart than broad at their bases. Petiole short, hardly Ih times as long as broad, sides of node convex when seen from above and much broader than the peduncle; in profile the node is high and very thick with an evenly rounded summit, very steep and concave anterior, and very abrupt posterior declivity; ven- tral tooth rather large, blunt . Postpetiole twice as broad as the petiole, distinctly broader than long, transversely elliptical from above, with rounded anterior angles. Gaster of the usual shape. Mandibles and clypeus subopaque, the former longitudinally striated and with a few coarse punctures, the latter longitudinally rugose. Head opaque throughout, evenly and densely punctate except along the sides of the frontal region, where there are a few delicate longi- tudinal rugse. Thorax, petiole and postpetiole opaque, densely punctate. Gaster smooth and shining. Hairs white, moderately numerous, clavate on crown of head, thorax and abdomen ; erect on head and thorax, more reclinate on the pedicel and gaster; minute, inconspicuous and appressed on the an- tennae and legs. Whole body, even the anterior portion of the gaster, golden-yellow, except the antennal club which is blackened. Female (dealated). — Length 2.5 mm. Head opaque, densely punctate; cheeks and whole preocellar region irregularly longitudinally rugulose, the postocellar region and posterior angles more reticulate. Thorax opaque, densely punctate ; mesonotum and scutellum wath very faint, parallel, longitudinal rugse. Epinotal spines very short and stout, regularly conical, not as far apart as they are broad at the base. Petiole and postpetiole like those of the worker, but the latter segment fully twice as broad as long. Both segments of the pedicel densely punctate, opaque, except the dorsal surfaces of the nodes which are somewhat smooth and shining. Body brownish- yellow, legs pale-yellow. Antennal club, wing-insertions, sides and posterior border of first gastric segment, a broad transverse band on the second gastric segment and the tip of the gaster, dark-brown. Pilosity like that of the worker. Type locality: Austin and McNeil (Travis county), Tex. This small species lives in and on the ground. At Austin I have occasionally seen a few workers running about on the dry gravelly hill-slopes exposed to the sun. At McNeil I took a few dozen workers and a dealated queen, which were inhabiting a small spherical chan:ber 256 PROCEEDINGS OF THE ACADEMY OF [Feb., in the "black waxy" soil under the center of a large flat stone. Two other nests taken at Austin exhibited a tendency toward plesiobiosis. One of these was found in the vegetable debris at the very entrance of the nest of a timid fungus-growing ant {Tracliymyrmcx turrifex Wheeler), the other under a stone at the very edge of a flourishing colony of Pheidole instabilis Emery. L. terrigena is undoubtedly closely allied to L. Andrei Emery, which I have not seen. The worker differs, however, in its decidedly smaller size, shape of petiole and postpetiole, its larger and stouter epinotal spines, the absence of a clypeal sinuosity, a shining longitudinal band on the front and vertex, and the absence of infuscation on the abdomen, although the last character is clearly present in the female. 18. Leptothorax Andrei Emerj-. L. Andrei Emery, Morpli. Jahrb. Abth. f. Syst., VIII, 1894, pp. 318, 322. Worker (PI. XII, fig. 22).— Length 2.25 mm. Clypeus feebly carinate in the middle, its anterior border subsinuate. Antennae 12-jointed; first funicular joint a little shorter than the three succeeding joints; second joint of club a little longer than the preceding joint. Thoracic dorsum without a mesoepinotal constriction. Epi- notal spines in the form of short, stout teeth. Petiole with a rather long peduncle, its node above subrotund. Postpetiole about a third broader than the petiole, a little broader than long. Mandibles striated. Head opaque, longitudinally rugulose-punc- tate; cheeks and clypeus striated, the latter and a median line along the front and vertex shining. Thorax and pedicel opaque; the former densely, the latter more faintly punctate. Gaster and legs shining. Hairs on the body sparse, short and clavate. Testaceous, abdomen darker behind, legs pale. Type locality: Cahfornia (Andre). The species was described from a single specimen in the collection of Prof. Emery. 19. Leptothorax (Dicliothorax) Pergandei Emery. L (£>.) Pergandei Emery, Zool. Jahrb. Abth. f. Svst., VIII, 1894, pp. 318-323, 324. ? . Worker (PI. XII, figs. 23 and 23a).— Length 2.5-3.25 mm. Mandibles rather broad, 5-toothed. Clypeus moderately convex, broadly rounded in front, with a distinct median carina. Antennae 12-jointed, scape extending beyond the posterior angle of the head a distance fully equal to twice its breadth; first funicular joint as long as the three succeeding joints together; joints 3-8 nearly as long as broad; club 3-jointed, the two basal joints subequal, together shorter 1903.] NATURAL SCIENCES OF PHILADELPHIA. 257 than the terminal joint. Thorax long, rather robust, not much wider in front than behind, without abrupt decUvity at the juncture of the neck and pronotum, pro- and mesonotum convex; mesoepinotal con- striction very deep and broad. Epinotal spinas small, not longer than broad at the base, directed upward. Petiole from above nearly three times as long as its greatest width which is in the middle ; in profile the node is low and rounded above, the anterior slope very long, at first nearly horizontal, then gradually ascending, the posterior slope shorter, somewhat flattened; summit of node distinctly impressed or concave when seen from behind; ventral surface of peduncle with a long but not very prominent tooth. Postpetiole fully half again as broad as the petiole, a little broader than long, nearly square, its anterior angles prominent, its dorsal surface in profile very convex, especially in front. Gaster rather large, of the usual shape. Sting well developed. Legs robust. Mandibles longitudinally striated. Clypeus smooth, especially be- hind, its sides longitudinally rugose. Head smooth and shining above and behind, sparsely punctate and irregularly and delicately reticulate. Antennal foveae with curved, parallel ruga%- front and cheeks with straight rugae. Neck opaque and delicately rugose; pro- and meso- notum shining, very finely and irregularly reticulate, pleurae, epinotum and mesoepinotal constriction subopaque, coarsely reticulate-rugose; in the constriction and on the meso- and metapleurae the rugae have a distinctly longitudinal trend. Declivous surface of epinotum smooth and shining. Petiole and postpetiole shining and finely reticulate above, more opaque and reticulate-rugose on the sides. Gaster smooth and shining, finely and irregularly reticulate. Hairs white, long and very abundant, obtuse but not clavate, erect on the trunk, suberect on the legs and antennae. Black or rich dark-brown, mandibles, antennae, except the club, frontal carinae, thorax, pedicel and legs yellowish-red or testaceous. In some specimens the dorsal surface of the thorax and nodes is blackened, while in others the whole of the thorax, nodes and legs is black except the peduncle of the petiole and the tarsi and joints of the legs, which are yellow. Female. — Length 3.5-4 mm. Apart from the usual sexual characters, differs from the worker in having the lower surfaces of the head, thorax, pedicel and the legs more yellowish. The smooth surfaces of the body are hardly reticulate. Mesonotum adorned with a median brown blotch on its anterior half and a large comma-shaped spot on either parapsis. Scutellmii, epi- 17 258 PROCEEDINGS OF THE ACADEMY OF [Feb., notum and pleurae more or less spotted with brown, and a small black spot at the insertion of the forewing. Femora and tibise infiiscated in the middle. Antennal club infuscated. Wings milky-hyaline, with colorless veins and stigma. Concavity at summit of petiolar node very distinct; postpetiole nearly twice as broad as long. Epinotal spines distinctly shorter than broad at their bases. Male. — Length 2-2.25 mm. Head a little longer than broad; cheeks very short, eyes and ocelli prominent. Mandibles overlapping, 4-toothed. Clypeus convex. Antennae 13-jointed, scape about as long as the five succeeding joints, first funicular joint about as long as the second and third together; joints 3-8 slender, cylindrical, subeciual, twice as long as Inroad ; club 4-jointed, the three basal joints subecjual, together as long as the terminal joint. Thorax long, mesonotum rounded, projecting forward, so that the head is scarcely visible when the insect is viewed from above. Epinotum with a pair of slight projections in the place of the spines. Petiole long and slender, like that of the worker luit with lower node. Postpetiole as long as broad, square from above, in pro- file with the node highest in the middle of the segment and rounded. Gaster and legs of the usual conformation. Mandibles and clypeus subopaque, the latter coarsely rugose. Head shining, very irregularly and sparsely reticulate, in front of the ocelli with a few shallow foveolse. Pronotum subopaque, reticulate; meso- notum very smooth and shining, finely reticulate. Scutellum and epinotiun more opaque, disk of former coarsely reticulate, sides of both longitudinally rugulose. Petiole and postpetiole opaque, more shining above, delicately corrugated. Gaster very smooth and shining. Hairs white, rather sparse and long, even on the legs and antennal scape, where they are reclinate but not appressed. Black. Mandibles, antennae and legs white except the following portions, which are infuscated or blackened: Edges and teeth of man- dibles, terminal joint of antennae, basal two-thirds of coxae, middle of femora and tibiae and last tarsal joint. Wings like those of the female. Type locality: Washington, D. C. (Pergande). Additional localities : Morgantown, N. C. (Forel) ; Austin, Tex. ; San Angelo, Tom Green County, Tex. ; Toronto, Brewster County, Tex. The type specimens of this fine species were taken by Mr. Pergande in a nest of Monomorium minutum Mayr var. minimum, and it was supposed that the Leptothorax was a guest in the nests of the Mono- morium, but Forel, who observed the species in North Carolina, showed that this was altogether an exceptional case. He found L. Pergandei 1903,] NATURAL SCIENCES OF PHILADELPHIA. 259 living "in independent formicaries, in the moss of woods or in the earth of meadows, hke the ordinary species of Leptothorax" (Aim. Soc. Ent. de Belg., Tome XLV, 1901, pp. 389-398). In Texas I have had ample opportunity to observe the habits of this ant, especially in the neigh- borhood of Austin, where it is found making its nests in very sparsely grassy spots among the mesquite and Opuntia thickets. The nests can be found only by carefully tracking foraging workers, as the entrance is a small hole often concealed under a dead twig or a tuft of grass roots. The colonies are hardly more populous than those of other species of Leptothorax. The winged forms appear during the last week in April and the first week in May. The workers run about on the soil in the hot sun as fierce hunters of small insects (Aphids, minute Heteroptera, etc). As they are extremely pugnacious even toward individuals of the same species from other nests, and as I have never found them nesting with Monomorium jninimum, though this species is very common in the same localities, I believe, with Forel, that Pergande's observation must be cjuite exceptional or may even involve some misinterpretation. 20. Leptothorax (Dichothorax) floridanus Emery. L. (D.) floridanus Emerv, Zool. Jahrb., Abth. f. Syst., VIII, 1894, pp. 318, 324. ^ . According to Emery, the worker of this species (PI. XII, figs. 24 and 24a) differs from the preceding in the following characters : The body is more shining, the epinotum smooth and shining above, the mesoepinotal constriction punctulate, subopaque, the petiolar node is narrower, and not impressed above, the postpetiole is hardly ^ again as broad as the petiole and proportionally narrower than in Pergandei. Type locahty: Florida (Pergande). Additional locality: North Carohna (Forel). The differences between the two Dichothorax are so slight that Emery suspected floridanus to be merely a subspecies of Pergandei. I am myself strongly of this opinion, but as I have seen only a single specimen of floridanus, kindly given me by Prof. Forel, I hesitate to reduce this form to subgeneric rank. In my specimen the petiolar node is very decidedly convex when seen from behind, and the epinotal spines are longer and more curved than in any of my specimens of Pergandei. In other respects I can see no differences of importance. Color, pilosity and sculpture are the same in both forms. 260 PROCEEDINGS OF THE ACADEMY OF [Feb., Explanation of Plate XII. Fig. 1. — Leptothorax muscorum Nylander. Switzerland. In this and the fol- lowing figures only the profile view of thorax and pedicel (petiole and postpetiole) of the worker are represented. All the figures are from camera lucida drawings. Fig. 2. — L. muscorum var. sordidus var. nov. Fig. 3. — L. acervorum Mayr. Switzerland. Fig. 4. — L. acervorum subsp. Canadensis Provancher. Fig. 5. — L. acervorum subsp. Canadensis var. Yankee Emery Fig. 6. — L. Emersoni Wlieeler. Fig. 7. — L. Schaumi Roger. Fig. 8. — L. fortinodis Mayr. Fig. 9. — L. longispinosus Roger. Fig. 10. — L. curvispinosus Mayr. • Fig. 11. — L. curvispinosus subsp. ambiguus Emery. Fig. 12. — L. curvispinosus subsp. rugatulus Emery. Fig. 13. — L. curvispinosus subsp. annectens subsp. nov. Fig. 14. — L. Schmittii sp. nov. Fig. 15. — L. nitens Emery. Fig. 16. — L. texanus sp. nov. Fig. 17. — L. fricarinatus Emery. (After Emery.) Fig. 18. — L. neomexicanus sp. nov. Fig 19. — L. obturator sp. nov. Fig. 20. — L. nevadensis sp. nov. Fig. 21. — L. terrigena sp. nov. Fig. 22. — L. Andrei Emery. (After Emery.) Fig. 23. — L. (Dichothorax) Pergandei Emery. 23a node of petiole seen from behind. Fig, 24. — L. {Dichothorax) floridanus Emery. 24a node of petiole seen from beh'nd. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 261 NOTES ON THE CONRAD COLLECTION OF VICKSBURG FOSSILS, WITH DESCRIPTIONS OF NEW SPECIES. BY THOMAS L. CASEY. The following paper anticipates a more elaborate descriptive cata- logue of the Vicksburg fossils which the author has in view, and it is intended that the species here described as new shall be figured with the others when that is published. It will be sufficient to state that nearly all the species here named for the first time are well known, and represented in every collection from Vicksburg, Red Bluff or Jackson, that can be considered in any wa}'' complete, and there can be but little doubt of their ready identification from the diagnoses, comparative or otherwise, which it is thought are drawn up with sufficient fullness and accuracy for that purpose. PELECYPODA. Corbula laqueata u- sp. C. filosa Con. (Am. Jour. Cone, I, p. 145; nomen prseoc). Differs from the species named perdvbia by De Gregorio; at least on comparing Vicksburg specimens witli those from Red Bluff, I find that the latter form is shorter and differs greatly in sculpture in the umbonal region. The true rugse do not begin for a considerable dis- tance from the beak in the Red Bluff form, this region being smooth or feebly wrinkled, whereas in the Vicksburg species the rugae begin very near the beaks, and the latter are more strongly gyrate. As the Vicksburg species is distinct I would propose the name given to it in manuscript by Conrad. C. laqueata is a small species, never materi- ally exceeding 6 mm. in length by 5 in height, the posterior flattened surface feebly delimited, the bounding line being rounded and not cari- nate ; the rugse are comparatively coarse ventrally. It is confined to the upper marl and represented in the lower limestone by a variety having much finer rugae. The type of Corbula interstriata appears to be the only known repre- sentative of the species, and I have never been able to discover it at Vicksburg. I did not examine the hinge with much care, but the general appearance of the species is not very unlike that of C. gibbosa 262 PROCEEDINGS OF THE ACADEMY OF [Feb., of Lea, being shorter and relatively higher, and it is probably a true Corbula. Under his description of Corbula aliforrnis {Am. Journ. Cone, II, p. 76), Conrad remarks that "this shell belongs exclusively to the Shell Bluff group and is very distinct from C. alia of the Mcksburg group." This statement is difficult to comprehend as C. aliformis has been found in abundance by Mr. C. W. Johnson in the bluff bordering Mint Spring Bayou, at Vicksburg, at a point only a few feet removed in elevation or horizontal distance from the pocket of sandy clay in which I have taken the true C. alta plentifully. It is presumable, therefore, that both these species existed contemporaneously in the Lower Vicksburg. They are both completely unknown from the Upper Vicksburg marl. There is no reason to suppose that other species of this same subgenus of Corbula (Tiza De Greg.) may not exist at Vicksburg, and in fact I have a single valve that seems to indicate a third species, much more equilateral than the others. The small Mactra occurring in great abundance in the Lower Vicks- burg is a different species from funerata, which occurs only in the upper marl. It is probable that the former, which is much more inequilateral, may be the one named imequilateralis by Meyer (Bull. 1, Geol. Surv. Ala., p. 82), although the figure is rather poor. It is sin- gular that the corresponding species occurring in the Jackson is the counterpart of funerata from the Upper Vicksburg, and differs de- cidedly from the Lower Vicksburg species. In the Conrad Catalogue {Am. Journ. Cone, 1865) there are two species which appear to have been originally named Psamrnobia missis- sippiensis, one under the genus Gari, on page 4, the other under the genus Abra, on page 5; the references seem to show that they were separately described and figured. The species Abra mississippiensis is the only one of which I can find the type. It is broadly oval, but slightly inequilateral and of moderately large size, relatively higher in form than Abra perovata, with which it occurs very abundantly in some parts of the Lower Vicksburg. The Gari mississippiensis {I.e., p. 4) I cannot place and there appears to be no type in the Conrad collection. There seems also . to be no type of Tellina perovata Conrad, and I have not been able to identify this species from the material in my cabinet. In the list referred to, the locahty "Claiborne" is attached to this species, but probably in error. The Abra protexta of Conrad, of which also no type can be found, is very abundant in the Lower Vicksburg, to which it is entirely confined; but it is a Tellina 1903.] NATURAL SCIENCES OF PHILADELPHIA. 263 near vicksburgensis, if I have identified it correctly from the published drawing, the latter being about twice as large as any of my specimens, but agreeing perfectly otherwise. Tellina pilsbryi n. sp. There is an undescribed Tellina occurring in large numbers in the Lower Vicksburg, to which I take pleasure in giving the above name in honor of Dr. H. A. Pilsbry, of Philadelphia. It is strongly inequi- lateral, trigonal, with the ventral edge rounded, rather thick in sub- stance, compressed, the anterior and posterior sides broadly rounded, the surface equally declivous and convex toward the edges and not more abruptly declivous along the anterior edge ; it is strongly marked with fine, very close-set concentric striae throughout. The lateral teeth of the left valve are large and strongly developed, those of the right obsolescent. The larger cardinal of the left valve is subbifid. The length of a nearly full-grown valve is 9 mm.; height about 7 mm.; beak about 2 mm. in front of the median line. The lunule is small and slender. Lucina vicksburgensis n. sp. Occurs in the Lower Vicksburg in great abundance. This species is subec|uilateral, suborbicular, rather compressed, the beaks somewhat high and acute, the adjacent anterior sinus small and deep, the lunule small and deep-set and but little more than twice as long as high as a rule. The surface is covered throughout with strong concentric raised lines which are close-set and low, but becoming strongly lamelliform on the anterior and posterior dorsal declivities, the dorsal edge being ren- dered rough spiculose and uneven thereby. Cardinal tooth of right valve very oblique. Length 7^ mm. ; height 7^ mm. It grew some- what larger than the type above described, but never exceeded this length by more than about 2 mm. It might be considered closely allied to the Jacksonian Cijdas curtus of Conrad {Am. Journ. Cone, I, p. 139), as the posterior side is noticeably more truncate than the anterior, were it not for the fact that curtus is described as "ventricose," a term which could not possibly be applied to vicksburgensis. Lucina scopularis n. sp. Red Bluff formation. Approaches vicksburgensis very closely, being orbicular and only moderately convex, but it is a little smaller and the concentric lines are finer, relatively less close-set and more lamelliform, becoming still more strongly so on the anterior and posterior dorsal declivities, though relatively less strongly so than on the corresponding parts of vicksburgensis. It differs principally 264 PROCEEDINGS OF THE ACADEMY OF [Feb., from the latter in the form of the anterostral sinus, which is longer and more transverse in outline, in the form of the lunule, which is larger and very much more elongate, and in the cardinal tooth of the right valve, which is here scarcely at all oblique, being almost perpendicular to the hinge line. In both species the laterals are moderately developed and the ventral edges smooth internally without trace of crenulation. Length of a moderately large specimen 7 mm., height 6.2 mm. The anterior and posterior sides are almost equally and very broadly rounded. This species is very shghtly more inflated than vicksburgensis. There is a variety which is very abundant in the Jacksonian of Moody's Branch. In the Conrad collection the only species of Lucina that I could dis- cover is the largest form occurring at Vicksburg — moderately convex, very thin in substance, with feeble hinge and nearly smooth surface. The label attached states that this is Lucina perlevis. This is, however, the species which w^as subsequently described under the name missis- sippiensis. There is some confusion here. Venericardia vicksburgensis n. sp. The larger Venericardia of the Vicksburg differs specifically from the one occurring so abundantly in the Jackson — named diversidentata by Meyer — and also rotunda of the Claiborne, in having very much fewer radiating ribs, these being only 19-20 in number. It is found only in the lower hmestone, and attained a length of 22 mm. or more. It occurs also at Red Bluff. Cardita aldrichi n. sp. Lower limestone at Vicksburg occurring in considerable abundance. It is only moderately convex, inequilateral, with about 15 broad, feebly convex, approximate radiating ribs. The longerribs, behind the middle, are more notably wide and separated throughout their length by scarcely half of their own width. The anterior lateral hinge tooth is distinct. The length of a specimen rather more than half grown is 5.5 mm., the height 4.5 mm. Area invidiosa n. sp. From the Red Bluff formation of Mississippi I have before me speci- mens of a small Area, probably allied somewhat to the Claibornian rliom- hoidella of Lea. It is subrhomboidal, very inequilateral, moderately inflated, broadly rounded ventrally, the anterior and posterior sides oblique, the former rounded, the latter longer and nearly straight. The beaks are rather broad, moderately elevated above the hinge-line, bisected by a feeble depression which becomes obsolete ventrally. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 265 The hinge-line is straight externally, broadly feebly arcuate internally, the line of teeth more than three-fourths as long as the shell, the lateral teeth becoming longer and strongly oblique. The space between the beaks and the hinge-line in flattened, nearly smooth except some fine, close-set parallel hnes of growth, but at the posterior end there are some coarser parallel and feebly oblique lines. The radial ribs are 28 to 31 in number, rather coarse and separated by much less than their own widths, except in the feebly depressed area radiating from the middle of the beaks where they become finer and relatively much more widely separated, and generally with one fine intermediate rib between them in this region toward the ventral margin only ; the ribs also become smaller but very close-set posteriorly in the flattened area toward the hinge-line. The surface posteriorly at an angle of about 30 degrees with the hinge-line is convex, becoming rapidly declivous and explanate to the latter. The muscular scars are rather deep. Lines of growth produce feeble transverse and rather widely separated nodules on the ribs generally becoming obsolete posterior^. The length of a moderately large individual of this species is 11.5 mm., the height 6 mm. Area delicatula n. sp. Occurs in the Lower Vicksburg limestone in great abundance. It may be regarded as a homologue of invidiosa and is doubtless one of the smallest known members of the family. It is elongate, very inequilateral, obliquely parallelogramic, moderately inflated, becoming flattened posteriorly toward the hinge-line, the latter long, thin and straight, the teeth small. The space between the hinge-line and the beaks rather low, flat and smooth or nearly so, narrowing very gradually posteriorly. The umbonal impression, with its diminished ribs, is nearly as in invidiosa and many other species. The ribs are some 28 in number, relatively moderately coarse, being generally sepa- rated by nearly their own wddths, flattened. Length of a moderately large valve 6 mm., height 2.6-2.8 mm. It is somewhat singular that no reference has been made to the very different sculpture of the right and left valves of Area lesueuri Dall (mississippiensis Con.). The left valve has the diverging ribs double, the pairs being much more close-set than the single and smaller ribs of the right valve. It results from this that the left valve is much the stronger and more frequently preserved intact. Area vaughani n. sp. While mentioning the genus Area, it may be appropriate to allude to a species, quite common in the Lower Claiborne at St. Maurice, 266 PROCEEDINGS OF THE ACADEMY OF [Feb., La., and allied somewhat to rhomboidella Lea. It attained a length of more than 20 mm., with a height of 12 mm. or more, obliquely rhomboidal, moderately inequilateral, rounded anteriorly and posteri- orly and broadly rounded ventrally. It is moderately inflated, the radiating concavity at the middle of the umbones almost obsolete and having merely shghtly wider intervals between the ribs, the latter 41-43 in number. The hinge-line is long and straight, the teeth becom- ing larger and very oblique laterally but well developed throughout, with their sides finely ribbed, giving to each tooth a bipectinate appear- ance. The area under the beaks is ample and broadly, divaricately striate. This species differs from rhomboidella in its much larger size, more numerous ribs, rounded ventral edge and many other characters, and may be named vaughani. A fair illustration of it was given by Mr. Vaughan (Bull. Geol. Surv., 142, PI. Ill, fig. 8), in whose honor it is named. A modification of the true rhomboidella, but still smaller in size, also occurs sparingly at St. Maurice. SCAPHOPODA. Dentalium strenuum n. sp. In the Upper Vicksburg there are two large species of Dentalium; one — jD. mississippiefise of Conrad — is moderately large, gradually tapering throughout its length, feebly, evenly arcuate, having about 12 well-marked raised threads which l^ecome doubled or sometimes quadrupled in number anteriorly, but generally almost effaced at the mouth. A moderately large specimen measures 47 mm. in length by 4.6 mm. in maximum diameter. The other species, which may be named strenuum,, is much larger, nearly straight, but becoming more rapidly arcuate and also more distinctly tapering in form near the posterior end. The ribs are some ten in number at the smaller end, becoming generally quadrupled in number at the mouth, where they still remain very distinct. The substance of the shell is much thicker, being fre- quently 1.2 mm. through the walls near the middle. The notch at the smaller end is nearly as in mississippiense, but generally deeper and more acute. The largest entire specimen in my cabinet measures 67 mm. in length by 6.3 mm. in maximum diameter, but I have seen frag- ments measuring more than 7 mm. in diameter and which represented examples probably not much less than 90 mm. in length. Dentalium opaculum h. sp. Occurs in the; Lower Vicksburg in very great numbers. It is smaller than mississippiense, somewhat less arcuate, gradually taper- ing, notably uneven in growth and frequently more or less contorted at 1903.] NATURAL SCIENCES OF PHILADELPHIA. 267 various points in its'extent, smooth but dull in luster, devoid of any trace of elevated ribs or threads except toward the smaller end, where some 12 to 16 faintly raised subequal hnes become visible. The pos- terior notch is very feeble and broadly angulate, much feebler than in either of the preceding species. A moderately large example meas- ures 40 mm. in length by 4 mm. in diameter, but the latter dimension occasionally attains 4.5 mm., which would represent a rather large individual. Dentalium zephyrinum ". sp. The commonest species at Red Bluff, closely resembUng the preceding in general size, form and slight irregularity of growth, but the longi- tudinal threads are distinct throughout the length and of a different form, being wider and flat, equal, about 16 in number, very strong posteriorly, becoming finer and feebler anteriorly where one or two feebler intermediate threads become visible. The posterior notch is well marked, not broadly angulate but generally rather deeper than wide. The length of the largest individual before me is 41 mm., with a maximum diameter of 4.2 mm. Dentalium polygonum n. sp. This species also, from Red Bluff, is still more slender, and is peculiar in being a perfect heptagon in cross-section near the smaller end, the angles of the polygon being minutely elevated, form- ing fine but conspicuous longitudinal threads, which remain distinct to the larger end; the intervals soon acquire two to four finer threads which never become as conspicuous as the primary ribs. The notch is not present on the truncated apex of the only specimen before me. Length 33 mm., width 3.2 mm. GASTROPODA. The Fusus jnississipicnsis of Conrad is a Latirus allied to jyrotractus, having the columellar Jolds evident though rather feeble. It differs in the entirely rounded outline of the whorls, there being no wide double band or collar below the suture as in that species. It is moderately abundant in the upper marls. Fusus vicksburgensis is very rare and alsf) occurs solely in the upper marls. My specimens are all fragmentary, partially decorticated and decomposed. Its broad flat lyra easily distinguish it and the nucleus and nepionic whorls are also peculiar. The Vicksburg type of Clavella differs from that of the Eocene horizons in having a small conoidal nucleus. They are probably generically different. As far as I have been able to discover there are 268 PKOCEEDINGS OF THE ACADEMY OF [Feb., at least two species in the Vicksburg, both confined to the lower hori- zon. There are several species of Pleurotoma in the Vicksbiirg and sub- adjacent horizons allied more or less closely to servata Con., and appar- ently neglected or overlooked hitherto in our literature, but which can be distinguished readily by brief comparative descriptions. In servata the nucleus is smooth, elevated, acute and of 3 or 4 whorls, the last whorl gradually acquiring a few riblets which become by degrees the 7 or 8 large rounded ribs of the body whorls. There is no appearance of a subcentral revolving carina on the whorl adjoining the nucleus, and on each of the more recent whorls there are generally 3 coarse revolving lines thickened on the ribs, with numerous very fine close-set intermediate threads, all occupying about lower half of the whorl. Just below the suture there is a conspicuous thickened collar, immediately below which there is a deep revolving concavity, the surface thence expanding to the uppermost of the coarse revolving lines, the entire surface between the latter and the collar having fine subequal and rather close-set lines. The aperture and canal together constitute about three-sevenths of the total length of the shell. Pleurotoma vicksburgensis n. sp. This species occurs plentifully in the Vicksburgian beds, accompany- ing servata and generally confounded with it. It usually attained a a little larger size and stouter form, and may be distinguished at once by the fact that the whorl immediately adjoining the nucleus has a strong revolving line below the middle, thickened on the ribs and accompanied by a close-set smaller revolving line immediately above it. The larger whorls generally acquire two other coarse, though much smaller revolving lines, one above and one below the two mentioned, and also finer intermediate threads. Just below the suture the elevated collar is not quite so prominent as in servata, and, instead of the abrupt concavity adjoining, the surface is almost evenly con- cave and rapidly expanded to the system of coarse revolving lines referred to, this surface being also finely, evenly lyrate. The nucleus is much shorter than in servata, consisting of between two and three whorls, and is not higher than wide. The aperture and canal are nearly as in servata. One of the larger specimens before me measiu^es 27 mm. in length by 7 mm. in width. The double carina of the nepi- onic whorls remains throughout the most conspicuous feature of the revolving sculpture, the lines becoming gradually more nearly equal and more widely spaced, with the dilatations on the ribs much more pronounced than in servata; the ribs, also, are much more broadly 1903.] NATURAL SCIENCES OF PHILADELPHIA. 269 rounded than in that species and become obsolete in the posterior concave area of the whorls. A specimen in the Conrad collection is marked "servata var.?"; it is deprived of the nucleus and adjoining whorls. Both this species and servata occur also in the Lower Vicks- burg, but in slightly modified forms. Pleurotoma oblivia n. sp. This is a Red Bluff species somewhat allied to servata. It resembles servata in general form and conformation of the nucleus, nepionic whorls, aperture and canal, the elevated smooth nucleus of three or four whorls and nepionic spire whorls being formed in the same way, but the ribs are narrower, more strongly rounded, much more elevated and only about six in number, strongly marked throughout the length of the whorl and only becoming extinct at the rather fine sinuous collar just below the suture. The revolving sculpture consists of eight or nine coarse lyriE, more dilated on the ribs, the first three less coarse and subequal, those in anterior two-thirds of the whorl generally with one fine thread intermediate. Length 22 mm., width 6.5 mm. It may be readily distinguished from servata by the fewer, narrower and more elevated ribs, more equal revolving lines through- out the length of the whorls and absence of any defined posterior flat- tened or concave area on the latter. Pleurotoma evanesoens n. sp. In the Jacksonian of the Kimbrel Beds, outcropping on the Red River a few miles below Montgomery, occurs another P^ewrotoma, rather closely resembling servata in general form, but widely distinct in sculp- ture and even more elongate in form. The nucleus is smooth, rather higher than wide and has about three whorls, the subsequent whorls mutually subsimilar, each having scarcely six large, though feebly elevated, oblique ribs. The collar below the suture is moderately wide but obtuse and low, not sinuous, and, immediately below it, the surface is feebly concave and moderately rapidly expanded to the middle, where each whorl is obtusely tumid and prominent, the sur- face thence gradually declivous anteriorly to the suture. The low feeble oblique ribs become wholly extinct in the concavity, occupying almost posterior half of each whorl. The sculpture is very fine and feeble, consisting of relatively broad but very feebly elevated, flat, revolving lyra, which are very close-set and subequal, mutually separated by a single very fine thread of similar character. The entire sculpture is so feeble that it is very apt to be entirely effaced by water wearing. Length of the largest of the three specimens 270 PROCEEDINGS OF THE ACADEMY OF [Feb., before me 32 mm., width 8 mm. On the part of the larger body whorls below the shoulder, the revolving lyrae become more widely separated, with three fine threads intermediate, as a rule, but the lyrae are always flattened and in more or less low relief. Pleurotoma Mlgardi n. sp. From the Jacksonian of Moody's Branch, I have two species which appear to have been confounded with servata, though differing radi- cally therefrom in the structure of the nucleus. One, named as above, is almost similar to servata in size, form and in the number and form of the slightly oblique rounded ribs, but has the raised revolving lines some eleven or twelve in number, subequal in size among themselves and becoming only slightly larger on the anterior parts of the whorl. The nucleus differs very radically from that of servata or ohlivia^ being small, obtuse and composed of only one and a half to two whorls. The canal also is decidedly shorter. Length 16 mm., width 5 mm. Pleurotoma coUaris n. sp. This is the second species from Moody's Branch referred to under the preceding description. It is stouter, with a still shorter canal, the aperture and canal together constituting about two-fifths the entire length of the shell. The nucleus is small, obtuse and of about two whorls. Body whorls about seven in number, each with some seven or eight obtuse ribs and a wide and strongly elevated conspicuous collar just below the suture, the upper sm'face of the collar declivous to the suture and having two close-set revolving striae, tlie lower part acutely elevated. The surface below the collar is deeply concave, then rapidly expanding to the posterior of the three strong raised hues which occupy about anterior half of the whorl. The concavity is marked with many very fine close-set revolving lines and the spaces between the three large lyrae referred to also have each about three fine lines. Length 17 mm., width 6 mm. Pleurotoma arnica n. sp. Of the species allied to rotceelens and tenella, there are several forms in the Red Bluff stratum. One of these, named as above, is somewhat stout, sculptured nearly like rotoedens, excepting that the concave and rapidly expanded surface immediately below the sutural collar scarcely ever acquires more than about two fine threads which occupy its median parts, while in rota^dens there are numerous fine threads at this part of the larger whorls. The nucleus in eimica is larger than \^ rotwdens, and there are generally about three of its whorls covered with fine acutely raised riblets, instead of about two whorls, as 1903.] NATURAL SCIENCES OF PHILADELPHIA. 271 in rotoedens. The strongly elevated median revolving keel is similar to that of rotoedens, and double, but the nodules are coarser. Length 14.5 mm., width 4.5 mm. The corresponding dimensions of an equally well-grown specimen (that is, of six body whorls) of rotoedens, from the Upper Vicksburg, are 11 by 2.8 mm. Arnica may be regarded as a probable ancestor of rotoedens. Pleurotoma ancilla n. sp. The archetype of tenella in the Red Bluff may be thus named. It is nearly similar in form to tenella but smaller, the nucleus large and well developed, of nearly five whorls, approximately the last two having nmnerous fine acute riblets; it is higher than wide and acute. The subsequent whorls have a broadly obtuse revolving prominence just below the middle, which is closely ribbed, the ribs longitudinal and rounded; collar below the suture consisting of two approximate subequal and slightly uneven revolving lyrse ; space between the collar and median ribbed tumidity moderately expanding and having three or four fine subequal lines; just below the median tumidity there is a fine irregular line. Aperture and canal together short, scarcely more than a third the length of the shell. Length 13 mm., width 4 mm. The specimens measured has about six body whorls. Pleurotoma plutonica n. sp. Not rare in the Lower Vicksburg limestones. This species is rather slender, perfectly smooth and polished throughout, with scarcely a trace of revolving sculpture except on the beak, where there are some oblique widely spaced striae. The nucleus is smooth, acutely ogival, higher than wide and of about four whorls. The subsequent whorls have each about eight low rounded oblique ribs, which become obsolete in a revolving concavity below the suture. The first three, or thereabouts, of the body whorls have a rather pronounced, though obtusely rounded, swelling adjoining the suture beneath, but this is gradually lost on the larger whorls, these having but feeble traces of a raised band at the suture, the latter being a very fine, slightly sinuose and feebly impressed line. The canal is well differentiated from the aperture, and the two combined constitute about three-sevenths of the total length of the shell. Length 12 mm., width 3.7 mm. Another speci- men, represented by the spire alone, indicates that the species may attain a length of fully 15 mm. or more. There is no trace of this species in the upper marls. Pleurotoma intacta n. sp. Another species, equally well defined, may be named as above. It is small, moderately stout, fusiform, the aperture and canal, which 272 PROCEEDINGS OF THE ACADEMY OF [Feb., are not very strongly differentiated, together constituting nearly half the entire length of the shell. The nucleus is as wide as high, consisting of about three whorls, smooth but gradually acquiring the fine riblets which become the eight or nine rather narrow and subacutely elevated oblique ribs of the subsequent whorls, the latter short, about four in number in the largest specimen before me, the ribs angular in profile from base to apex of the whorl, with point of maximum elevation just below the middle of the length and becoming obsolete just below the pronounced uneven and closely duplex collar margining the suture beneath. Each whorl has six or seven coarse, subequal and closely approximate flattened or slightly convex lyrse, those in lower half slightly coarser than the posterior three, and that at the middle slightly thickened at the summits of the ribs. Length of the largest in an extended series 7 mm., width 2. ,5 mm. This species occurs only in the upper marls at Vicksburg and is common. PHANDELLA n. gen. This genus occurs, in the Upper Vicksburg marls, and appears to have no closely allied living descendant, although related perhaps to Daphnella. The shell is minute, and the animal apparently existed the greater part of its life in the nuclear stage, there being no example which I have seen, out of a considerable series collected, possessing more than ^between one and two body whorls. The nucleus is rela- tively large though evenly conical, pointed, consisting of from five to six whorls which are exquisitely sculptured in two systems of very minute lines crossing each other at an angle of about 45°, producing an appearance very much like the engine-turning frequently engraved upon a watch. There are before me three species,, distinguishable among them- selves by very clearly marked characters, but at the present time I will only briefly outline the most abundant of the three, which may be regarded as the type of the genus. Phandella nepionica n. sp. This species has about one and a half body whorls, which are together about twice as long as the nucleus, polished and completely devoid of revolving sculpture, having, however, about ten sharply elevated longitudinal or slightly oblique ribs, which become abruptly declivous posteriorly and obsolescent near the suture. The nucleus has about five whorls; the canal is rather short and there is a fine raised collar margining the suture beneath, which line may also be observed^ to 1903.] NATURAL SCIENCES OF PHILADELPHIA. 273 mutually separate the larger of the nuclear whorls. Length 2.25 mm., width 1.2 mm. Many specimens. Drillia harmonica i>. sp. A well-defined new species, quite rare in the Lower Vicksburg, and not yet found in the upper marls. It is rather stout, the spire apparently narrowing somewhat more rapidly toward apex. Nucleus simple and composed of three or four whorls. The sub- sequent whorls are rather short, each with some eight or nine strongly marked rounded ribs, longitudinal in direction or nearly so, and generally in line from one whorl to the next; they extend nearly throughout the length of the whorl, becoming obsolete only in the narrow revolving concavity below the ante-sutural elevated collar, which is rather thick and conspicuous and marked posteriorly with olie or two striae. Each whorl has some seven or eight nearly equal revolving lyrse, those near the middle mutually separated as a rule by a finer line. The aperture is rather wide, the canal very short, the two together but little more than a third of the total length, the callus near the posterior angle of the aperture tumid and conspicuous. Length 11 mm., width 3.7 mm. I had confounded this species with mississippiensis, of Conrad, until a recent inspection of the type of the latter shows that it is very different; mississippiensis is very stout much larger, with the revolving concavity below the sutural collar very wide, constituting about half the entire length of the whorl ; the short, broadly rounded ribs are confined to anterior half of the whorls and are obsolete in the posterior concavity. The specimen is somewhat water-worn, so that the sculpture is not distinct, but there are appar- ently revolving raised lines which distinguish the species at once from the smooth and otherwise very different ehoroides. The type seems to be unique. Scobinella pluriplicata n. sp. In the genus Scobinella, of Conrad, it should be stated that the species occurring at Red Bluff is distinct from ccelata of the Upper Vicksburg marls, and I would propose the above name for it. This species is much larger than coelata, with a relatively more elongate and less rapidly acuminate* spire, and differs also in sculpture. In cnelata there is a broad flattened duplex collar extending from the suture anteriorly for about a fifth the length of the whorl, the surface then concave to well below the middle, generally with about three revolving lines at the bottom of the concavity, the middle one of which is nodulose; the surface from the concavity to the lower limit of the 18 274 PROCEEDINGS OF THE ACADEMY OF [Feb., whorl is more prominent, flattened and divided into two coarsely nodose sections by a fine stria. In pluriplicata the whorls are rela- tively much more elongate, and, from the suture for about one-sixth of the length, are flattened; the next sixth of the length is occupied by a small concavity containing a nodulose line, which is even more prom- inent than the preceding flattened collar; the surface thence to the anterior limit of the whorl, occupying fully two-thirds of the length, is still more elevated but flattened, cj'lindrical and divided into about four nodose rings by three rather coarse equidistant revolving grooves. The canal is more prolonged and more obconic than in coelata, and the plications of the columella number some four to five. Length of a specimen of about seven body whorls 35 mm., width 9.5 mm. Length of a specimen of coelata of the same number of w^horls 21 mm., width 6.5 mm. Pluriplicata occurs also at Byram's Ferry. The Lower Vicksburgian at A^icksljurg has not yet yielded a trace of the genus. Scobinella famelica n. sp. Very slender and elongate, the aperture narrow, scarcely at all wdder than the canal, from which it is but feebly differentiated, both together constituting but little more than a third of the total length of the shell. The nucleus is rather small, of about three whorls, with its summit obtuse. Sul^sequent whorls each with a prominent double collar subjacent to the suture and a broad obtuse and strongly elevated revolving keel, fully a third as wide as the length of the whorl and divided into two subequal rings by a re- volving groove, situated below the middle of the whorl ; this duplex ring is obliquely and coarsely nodose. In the concavity between the collar and the elevated keel there are two or three fine revolving lines, the posterior of which is finely and more or less evenly nodulose. The anterior margin is a fine line on a level with the duplex ring and separated therefrom by a narrow deep concavity. Columella with six or seven rather unequal, close-set oblique folds, forming a slightly tumid columellar band as wide as the distance separating it from the posterior angle of the aperture. Length about 25 mm., width 4.5 mm. It occurs exclusively in the Upper Vicksburg marl and is rare. Scobinella maoer n. sp. Upper Vicksburg. This species resembles the preceding in general form and sculpture_.but has only two folds on the columella. These folds are strong, subequal and do not seem to be attended by any adventitious plicae. This species is elongate and very slender, the nucleus simple and of about three whorls. Each of the subsequent 1903.] NATURAL SCIENCES OF PHILADELPHIA. 275 whorls has a broad, moderately elevated double collar subjacent to the suture and a strongly elevated, obtuse and nodose double carina at a third of the length from the anterior margin, the deeply concave intermediate surface having a single strongly beaded line along the middle and a few other very faint and obscurely irregular revolving threads. The lower margin is moderately elevated, the surface thence to the large double carina concave. The spire before me consists of seven body whorls and is 10 mm. in length and about 3.5 mm. in width at base. The remainder of the shell is missing, it being very rare and represented thus far only by fragments. The genus Scobinella, of Conrad, is probably vaUd, but contains species having anywhere from two to six or seven columellar folds.^ They hold together very well in general type of sculpture and depart materially in a great many characters from Cordieria. The genus Cordieria does not occur in the American Eocene fauna and there is considerable confusion in the literature concerning it. Cossmann states that the first two of the species originally placed in Cordieria by Rouault are really Borsonia, and, assuming the third species as the type, gives an illustration of a "plesiotype" which would bear considerable resemblance to Latirus were it not for the obsolete canal. He also states that the embryo of Cordieria is paucispiral with subglo- bular apex, which does not agree with our species, such as hiconica, plicata and ludoviciana, and another genus may have to be made for these. In Scobinella the sinus is well developed, but in Cordieria and the American analogues named above it is very feeble, so that on this ground as well as many others the association of Cordieria and Scobinella by Tryon was entirely unjustifiable. In fact the Pleurotomid affinities of some of the forms assigned to Cordieria stand in. need of fuller confirmation. Turbinella perexilis, of Conrad, is a species having two strong columellar folds and is of uncertain relationship. We do not appear to have the true Borsonia in the American fauna, and Borsonia plenta, of Harris (Proc. Acad. Nat. Sci. Phila., 1895, p. 63), cannot properly be referred to that genus. Besides the species from Red Bluff and Vicksburg, referred above to Scobinella, this genus will include PL (Eucheilodon) reticulatoides Harris (I.e., p. 63), from the Lower Claiborne of Texas. The genus Eucheilodon, as represented by crenocarinatum of Heilprin, differs from Scobinella in the form of the nucleus, this being very large, probably indicating a slightly different line of descent. ^ If Pleurotoma (Moniliopsis) elaborata Con., be included, as I believe to be proper, the genus will have also species without columellar folds as is the case with the genus Microdrillia to be described below. 276 PROCEEDINGS OF THE ACADEMY OF [Feb., MICRODRILLIA n. gen. A number of minute Pleurotomids, including injans and cossmanni of Meyer, and harrisi of Aldrich, have been referred to by Cossmann under the namas Asthenotoma and Scobinella, by Harris under Mangilia, by Aldrich under Glyphostoma, and by Meyer, Vaughan and others under Pleurotoma in its broad sense. They are all very small and characterized by a well-developed, multispiral, closely coiled embryo, having one to three of its basal whorls costulate, few body whorls which are wholly devoid of costse but spirally carinate, the retral sinus relatively large, circularly rounded and close to the suture, the aperture oblique, columella callous, with or without plications, and the canal short or subobsolete. The genus Microdrillia differs from Asthenotoma, to which cossmanni was referred by Cossmann, in the structure of the embryo, and, espe- cially, in the position of the retral sinus, which in Asthenotoma corre- sponds in its greatest depth with the median line or periphery of the whorls. In fact, there is only one American species known to me which can properly be assigned to Asthenotoma, this being the PI. texana of Gabb. Microdrillia is much more closely related to Glypho- stoma, as suggested by Aldrich, but is not at all allied to Mangilia. It appears to have become wholly extinct in the Oligocene or Lower Miocene. The species were numerous and individually abundant, especially in the mid-Eocene of the Lower Claiborne, and those before me may be readily identified by the following table : Columella without folds, 2 Columella with numerous rather widely and evenly spaced folds; shell thick and heavy, the base angulate, not at all rostrate 8 2 — Shell rhomboidal in profile, thick and strong, the ante-peripheral part but little shorter than the entire portion behind the periphery of the body whorl; revolving carinae very thick 3 Shell more elongate, the ante-peripheral part always much shorter than the post-peripheral, thinner and more delicate, with relatively fine carina^. . 4 3 — Embrj'o small, evenly and broadly conical, of three smooth and one finely costulate whorls ; body whorls four in number, each with three strong carinsp, the lowermost carina first appearing generally on the second or third whorl; lines of growth strongly marked, cancellating the body whorl; columella subumbilicate in the type. Length 4.5 mm.; width 2 mm. Jacksonian Eocene of Moody's Brancli, Miss., . . . [Plewotoma] cossmanni Meyer [weycri Coss.] Embryo similar in structure I)ut much larger; body whorls generally not more than three in number, each with two very thick carinse; columella not um- bilicate in the specimens at hand; lines of growth very fine, scarcely notice- able. Length 2.8 mm., width L4 mm. Lower Claiborne Eocene. A^esey Creek, Lee Co., Tex solidula, n. sp. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 277 4 — -Peripheral carina at or above the middle of the wliorls ; concave f asciolar sur- face simple or with one or two fine simple revolving threads 5 Peripheral carina below the middle of the whorls; beak well defined and rapidly acuminate ; f asciolar surface with a conspicuously beaded thread ; embryo with two or tliree smooth and about two costulate whorls; body whorls three or four in number 7 5 — Base of the shell distinctly rostrate; body whorls about four in number. Form stout, the spire whorls short, about three times as wide as long, with a subduplex subsutural collar and three other single carinre, the fasciolar surface with two fine revolving threads; beak short and broad; lines of growth strongly marked; embryo well developed as usual, higher than wide, ovulate at tip, with three smooth and two costulate whorls. Length 5.7 mm., width 2.3 mm. Lignitic Eocene. Wood's Bluff horizon, rostrahda, n. sp. Base of the shell angularly acuminate, not at all rostrate; body wliorls variable, but never exceeding four as far as knowni 6 6 — Embryo as wide as high, obtusely oval at tip, with three smooth and two costulate whorls ; subsequent whorls four in number, each with one subsutural and three other equal carinte; fasciolar surface without a revolving thread. Length 5.6 mm., width 2.25 mm. Lower Claiborne Eocene. St. Maurice, La robustula, n. sp. Embryo acutely conical, scarcely higher than wide, with three smooth and one costulate whorls ; subsequent whorls not more than two in number in any of the three type specimens at hand, similar to those of robustula, but much more slender. Length 2.3 mm., width 1 mm. Lower Claiborne Eocene. St. Maurice, La minutissima, n. sp. Emljryo larger and much more complex, evenly conical, as wide as high, with two small smooth whorls and three finely and closely costulate, the latter strongly convex near their basal margins; form rather stout; whorls two in number in specimens at hand, the fasciolar surface without a revolving thread. Length 3.9 mm., width L5 mm. Red Bhiff Eocene. [Pleurotoma] infans Meyer Embryo nearly one-half higher than wide, subcylindrical, rapidly pointed at tip, with three smooth and nearly three coarsely costulate whorls, the latter strongly and more medially convex; subsequent whorls not exceeding four in number, the fasciolar surface with a fine revolving thread ; shell much more slender and elongate than in infcms. Length of embryo alone L4 mm., width .8 mm. Upper Vicksburg Oligocene vicksburgella, n. sp. 7 — Beaded thread below the middle of the fasciolar surface, a finer simple thread between it and the subsutural carinule; granulations of the beaded thread fine but distinct. Length of a specimen of 3.5 body whorls 6.7 mm., widtl^ 2.6 mm. Lower Claiborne Eocene. St. Maurice, La. [Pleurotoma] lerchi Vgn 8 — Spire more rapidly narrowed toward apex, the embryo very small, of three or four whorls, the lowermost apparently costulate; body whorls five to six in number, each with subsutural and submedian coarse equal carinse and a third, finer and less conspicuous, between the latter and the base ; concave fasciolar surface between the coarse carina with a fine median revolving thread; lines of growth well marked as usual; columella subumbilicate as a 278 PROCEEDINGS OF THE ACADEMY OF [Feb., rule. Length 6.5 mm., width 2.3 mm. Lower Claiborne Eocene. Moseley'.s Ferry, Burleson Co., Tex aldrichieUa, n. sp. Spire evenly, conically acuminate, the body whorls never exceeding four in num- ber; embryo always larger and well developed 9 9 — Revolving cariniB as in aldrichieUa, very coarse; fasciolar surface with a more or less distinct revolving thread; embryo as wide as high, conical, acutelv pointed, with three smooth and one costulate whorls; shell stout and thick. Length 5 mm., width 2 mm. Lower Claiborne Eocene. Elm Creek, Lee Co., Tex [Glyphostoma] harrisi Aid. Revolving carina? very much finer; shell narrower and more slender; embrj'o well developed, higher than wide, acutely conical, with four smooth and one costulate whorls ; fasciolar surface without a revolving thread ; columella with four or more folds as usual in this group. Length of specimen with three body whorls 4.5 mm., width 1.5 mm. Lower Claiborne Eocene. Elm and Vesey creeks, Lee Co., Tex elongatida, n.sp. Revolving carinse fine as in elongatula, the shell similarly slender, differing in having two costulate embryonic whorls, a fine revolving thread in the fasciolar surface and but two columellar folds; body whorls but two in number in the type. Length 2.3 mm., width 1 mm. Red Bluff Eocene. biplicatula, n. sp. A species which is strikingly similar to the Lower Claiborne harrisi was figured by Cossmann (Essais Pal. Comp., deux, liv., PL VI, fig, 35) under the name Scobinella Imvi'plicata Gabb. It is said to have been found at Jackson, Miss., by Meyer, but I have seen no plicate species from that horizon. Cossmann's generic diagnosis of Scobinella is also drawn from this figured "plesiotype," and for this reason does not apply to the Scohinella of Conrad at all. The true Iceviplicata is a Eucheilodon and is perfectly synonymous with reticulata Gabb, The reticulatoides of Harris is, however, a true Scobinella, the name being therefore somewhat unfortunate as Scobinella and Eucheilodon are amply distinct genera. Under the description of Pleurotoma infans (Geol. Surv. Ala., Bull. I, p. 75) Meyer refers to a Vicksburg form under the name var. brevis. This form is really not described at all and must be considered a list name, it being simply stated that it is decidedly stouter than infans. Even this statement, however, will not apply to the vicksburgella defined above, which is more slender and elongate than infans and differs radically in the form and sculpture of the embryo, Mr, Harris states {Bull. Am. Pal., Vol. 3, p, 24) that the Glyphostoma harrisi of Aldrich is a synonym of infans Meyer. This is incorrect, as subsequently held by j\Ir, Aldrich ; the two species are not at all closely related. 1903.] NATURAL SCIEXCES OF PHILADELPHIA. 279 COCHLESPIRELLA n. gen. The Fusus nanus, of Lea, belongs to a genus widely different from Microdrillia, but allied more closely to Cochlespira of Conrad. The genus, which will include also insignifica of Heilprin, which is not the same as nanus according to some very accurate drawings of the types very kindly lent me by Mr. Aldrich, and one or two undescribed Texan forms, may take the above name. Couus scopularis n. sp. In the Red Bluff deposit there is a Conus, about the size of the Vicksburgian alveatus, but with a much more depressed spire and differing also in sculpture to a marked degree. This species differs also from any of the forms found in the Jacksonian. It is rather broadly obconic in form, the exterior outhne of the body whorl straight, becoming very broadly, feebly arcuate toward the shoulder, just below which there are some two or three close-set and very obsolete obliquely rugulose revolving striae. The anterior oblique revolving striae are very feeble, subobsolete and occupy lower third of the whorl. The oblique anterior columellar fold is feeble. Spire extremely short and flattened, not more than an eighth or ninth as long as the body whorl, the whorls flat, each with five strong revolv- ing lyrae separated by equally large grooves, the arcuate lines of growth very distinct, producing a beaded or scabrous appearance. Length 28 mm., width 16.5 mm. The fine striae on the face of the body whorl at the top and parallel to the acute shoulder angle distinguish this spe- cies at once from any in the adjacent horizons. The species named Bursa mississippiensis by Conrad, in the list of 1865, is, in all probability, the young of Triton conradianus Aid., of the Red Bluff horizon. I have found this species in the Lower Vicks- burg. Bursa abbreviata is, however, a distinct species, moderately abundant in the upper marl at Vicksburg, to which it is confined, as is also Distorsio crassidens. Tritonopsis suhalveata, of Conrad, is confined to the Lower Vicksburgian, like Bursa mississippiensis. Phos macilentus n. sp. Phos mississippiensis, of Conrad, occurs in both horizons at Vicks- burg, but does not occur at Red Bluff. The species of the Red Bluff formation, which has hitherto been confounded wdth it, may take the name indicated. It is much narrower and more elongate than mississippiensis. The ribs are more numerous, finer and not so nodu- lose at the points where they are crossed by the revolving sculpture. The revolving lines are subequal among themselves, finer than in mis- 280 PROCEEDINGS OF THE ACADEMY OF [Feb., sissippiensis and not so strongly alternating in size. Length of a specimen of six body whorls 14 mm., width 4 mm. P. mississippiensis seldom has more than five body whorls, and an average specimen measures 13.5 mm. by 5 mm. Phos falsus n. sp. I have before me a remarkable Red Bluff Phos, which may be named falsus. It is very much larger than macilentus, though nearly as slender. The nucleus is as in that species and mississippiensis, consisting of four whorls, the lowest of which is sculptured with very fine obliquely sigmoid riblets. The body whorls are six in number, with rather widely spaced longitudinal ribs, some eight in number, subequal among themselves on the first four whorls, but then becoming very widely spaced and finally completely disappearing, leaving the surface even; the revolving lines are distinct but not very coarse, and are mutually separated on the larger whorls by two or three fine, closely spaced threads. The type before me has a strong rounded varix on the sixth whorl and another forming the outer lip. Length 19 mm., width 6 mm. Metula fastidiosa n. sp. In the Red Bluff bed there is an apparently undescribed Metula greatly resembling gracilis Johnson, from the Lower Claiborne of Texas {Proc. Acad. Nat. Sci. Phila., 1899, p. 75, PI. II, fig. 3). This species, which may be named as above, has a smaller and more rapidly pointed spire than gracilis, and has a greater number of varices. The nucleus is simple, smooth, rather higher than wide, ogivally pointed and of about three whorls, the subsequent whorls five in number, broadly, evenly rounded at the sides in profile, each with a feebly ele- vated fiattened varix, relatively rather wide, on which the longitudinal ribbing becomes obsolete and the revolving lyrse also obsolete except on the body whorl, where they continue uninterruptedly over the varix, which here becomes relatively still wider though so slightly elevated as to be scarcely definable. The ribs are small, and, from varix to varix on the spire whorls, about 32 in number; on these whorls the revolving grooves are about 10 in number, and, with the exception of the two posterior and one finer anterior, do not cross the ribs but appear as short excavated lines between them; on the body whorl, however, all the grooves cross the ribs but are reduced in width on their summits ; the ribs on the body whorl are also somewhat changed in character, being notably less steep in cross-section on the side lying in the direction of the growth of the shell. The columella is thickened 1903.] NATURAL SCIEN'CES OF PHILADELPHIA. 281 anteriorly below the middle and the aperture and canal together are half as long as the shell. Outer lip wdth a plicate band parallel with the edge at a short distance therefrom. Length 15.5 mm., width about 6 mm. Me tula fragilis n. sp. This is apparently the direct descendant of fastkUosa in the Upper Vicksburg marl and is a much larger species, with more numerous and relatively much finer and more close-set ribs; it is very thin and deli- cate in substance, very rare and always occurs in a fragmentary condition. From a fragment before me I am able to compute the diameter of the body whorl to be about 10 mm. The longitudinal riblets are at least 90 in number, and the revolving lines on the largest whorl of the spire about 20. No varices can be seen on the fragments before me. The revolving grooves are shallow, those near the base and apex of the whorls broader and more thoroughly olDliterating the ribs. The ribs are broadly arcuate longitudinally. The length of the specimen at hand must have been at least 25 mm. The shell walls are composed of three layers, of which the inner, very thin, and the outer, thicker, are solid and amorphous in texture, the two separated by a very thin layer of prismatic structure having the fibres perpen- dicular to the surface. Olivella affluens n. sp. There are two yery distinct species of Olivella occurring abund- antly in the \'icksburg strata. One of these, which may be assumed to be the typical 7nississippiensis , has the nucleus small, of very few whorls and the nucleal sutures obhterated. The whorls of the spire are completely unexcavated along the anterior edge. . The other species named as above, is as common as mississippiensis; it is rather smaller, the nucleus being, however, much larger, very obtuse, composed of about three whorls having the sutures all distinct and impressed. Each whorl of the spire has a deep and clearly defined revolving groove at the lower margin, which is entirely wanting in mississippiensis. The anterior folds of the columella are less oblique than in that species. Length of a moderately large specimen, having four body whorls beside the nucleus, 14.5 mm., width 5.7 mm. The aperture is somewhat nar- row, and is rather less than two-thirds the total length of the shell. The differences between these two species are at least subgeneric. It would seem to me that Fulgur spiniger and nochdatum, of Conrad, are distinct species and not mere varieties of one. I have collected a number of specimens of spiniger, and there is no marked variability 282 PROCEEDINGS OF THE ACADEMY OF [Feb., to be perceived. Spiniger is confined to the Upper Vicksbiirg marl, while nodulatum has occurred thus far only in the Lower Mcksburg limestone. The latter differs in both form and sculpture from spiniger, but varies of course in the relative amount of shoulder exposed above the suture, as is generally the case throughout the genus. The form occurring at Red Bluff differs from spiniger in its larger size, rather more elongated form, feebler sculpture, less strongly diffoi-en- tiated whorls of the spire, and more feebly developed spines of the spire whorls, also in usually having traces of a fine subobsolete, remotely and minutely subnodulose line at some distance below the spinose shoulder line. It may not be more than a subspecies of spiniger, but resembles that species more closely than nodulatum, from the intermediate beds constituting the Lower Vicksburg, a case very similar to that before referred to under Mactra. Lyria nestor n. sub-sp. The Lyria misssisippiensis, of Conrad, moderately abundant in the Upper Vicksburg marl, is represented in the Red Bluff bed b)^ this form which must be regarded as at least subspecifically distinct. It is much more elongate in outline than the Vicksburg species, and is more distinctly sculptured. The longitudinal ribbing is more obtusely rounded and less distinct. The length of a moderately large specimen is about 43 mm., with a maximum width of 18 mm. It is accurately figured by Dall (Trans. Wag. Inst., Ill, PI. 6) under the name Lyria costata Sowerby, and, in his opinion, both this and the Vicksburg form are varieties of that European species. The Vicks- burg and Red Bluff forms are, however, distinctly differentiated in facies and each holds to its own type through very extended series, without exhibiting much variability. Conrad's type of Conomitra staminca is apparently unique as far as the Vicksburg strata are concerned, and, from the matrix that partially envelops it, would appear to have come from the upper marl. It is a small species, about the size of the Claibornean fusoides and somewhat of the same form. The revolving grooves are deep and do not cross the ribs, thus forming short and very conspicuous excavated hues, nearly a^s in Fusimitra cellulifera Conrad. The species figured by Dr. Dall (I.e., PL 4, fig. 2) is quite evidently distinct from siaminea, and is a much larger species. It also seems to be distinct from angulata Heilp. The specimen figured by Dr. Dall has five body whorls and measures 15 mm. in length. A specimen of the true staminea, recently lent me by Mr. Aldrich and found at Byram's Ferry, having 1903.] NATURAL SCIENCES OF PHILADELPHIA. 283 four body whorls, measures only 8.5 mm. in extreme length. I have collected a number of specimens of M. vickshurgensis at Vicksburg, and feel sure that this is also distinct from staminea, having no suggestion of the peculiar deep coarse revolving sculpture of that species. The Mitra 7nississippiensis of Conrad is a specimen of conquisita in which the revolving lyration covers the entire body whorl; it is gen- erally effaced on the upper parts of this whorl in half-grown and older individuals, but occasionally persists until rather late in the growth of the shell. Millingtoni is a different species . The Capulus occurring at Vicksburg is apparently different from americanus of the Jacksonian, at least subspecifically. It is extremely rare at Vicksburg and has been found thus far only in the upper marl . 284 PROCEEDINGS OF THE ACADEMY OF [March, March 3. Mr. Arthur Erwin Brown, Vice-President, in the Chair. Seventeen persons present. A paper entitled "Studies in American ForficuHdse/' by James A. G. Rehn, was presented for publication. March 10. The President, Samuel G. Dixon, M.D., in the Chair. Seventeen persons present. March 17. The President, Samuel G. Dixon, M.D., in the Chair. Ten persons present. The death of Lewis Woolman, a member, on the 13th inst., was announced. Papers under the following titles were presented for publication: "The Mutation of Hibiscus moscheutos L.," by John W. Harsh- berger. "A Note on the Common Bottle-nose Porpoise of the North Atlantic, Tursiops truncatus (Montagu)," by Frederick W. True. "Additions to the Japanese Land Snail Fauna, No. YIII," by Henry A. Pilsbry. 1903.] natural sciences of philadelphia. 285 March 24. Edwin G. Conklin, Ph.D., Vice-President, in the Chair. Nineteen persons present. A paper entitled "Life Colors of Pcecilia limantouri and Description of a new Heros from Mexico," by Henry W. Fowler, was presented for pubUcation. The following minute on the death of Lewis Woolman was unani- mously adopted: The Academy of Natural Sciences of Philadelphia desires to place on record its appreciation of the loss it has sustained in the death of Lewis Woolman, who, since his election to membership in 1884, has manifested a constant interest in the well-being of the society. His work in connection with the geology of New Jersey and eastern Penn- sylvania, especially as illustrated by the boring of artesian wells, formed a valuable addition to knowledge. He was thorough and accurate in his methods, while his intercourse with his fellow-members was characterized by a hearty cheerfulness and sincerity. He was active in promoting the interests of the Biological as well as those of the Mineralogical and Geological Section, and during the latter months of his life he rendered service to the Academy as a member of the Commit- tee on Accounts. His memory will be held in grateful recollection. March 31. Charles Schaeffer, M.D., in the Chair. Ten persons present. The death of William V. McKean, a member, on the 29th inst., was announced. Charles Z. Try on was elected a member. The following were ordered to be printed: 286 PROCEEDINGS OF THE ACADEMY OF [March, THE VARIATIONS OF EUTSINIA IN THE PACIFIC SUBREGION, BY ARTHUR ERWIN BROWN, The portion of the Pacific coast of North America occupied by the garter snakes extends from about latitude 50° in British Columbia to the neighborhood of 33° in southern California/ and exhibits great variety of soil and climate, especially in the extremes of dryness and humidity. The annual rainfall at Puget Sound has reached a hundred and thirty inches, while at Yuma, in southeastern California, the aver- age is little more than three. Under these circumstances, and having in mind the ease with which color in reptiles is acted upon by external conditions, of which there is reason to believe that moisture is one of the most active, it is not surprising that color variation should reach a maximum in a group of snakes which, through diversity of habit, occupy practically every station open to their kind. Many species and subspecies have been established from this region upon examples so few in number as to form an altogether inadequate foundation in a genus where the range of variation is as great as it is known to be in Eutcenio — or is known, at least, to all but those naturalists who estabhsh uniformity by the easy process of attaching a name to every difference, without regard to its nature or its biological significance. Efforts to bring these supposed forms into some sort of order have been made in late years by Mr. G. A. Boulenger,^ Mr. John Van Den- burgh^ and the present writer,'* but it must be admitted that the assign- ment of some intermediates has been accomplished in part b}" the exer- cise of that mode of judgment which has been termed "the naturalist's instinct." Some material now in my possession bears directly upon the questions of identity involved in E. elegans and E. infernalis, and their interesting character has led me to review all the garter snakes of the region in question. The list following contains the names of all the species and subspe- cies within my knowledge, with the date of establishment and the ^ Eutwnia eques Reuss. has been found in the peninsula of Lower California, and others may yet be taken in tliat little-known region. ^ Catalogue of Snakes in British Museum, Vol. I, p. 192, et seq. (1893). ^ Occasional Papers California Academy of Sciences, V, p. -199, et seq. (1S97). * Proceedings Academy of Natural Sciences of Philadelphia, 1901, p. 18, et seq. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 287 type locality, which have been founded upon snakes belonging to the genus from this region, or whose range has been extended to enter it. Those admitted here are printed in capitals : Coluber parietalis Say (1823). Missouri river near Council Bluffs. Coluber infernalis Blain. (1835). California. Tropidonotus ordinoides B. and G. (1852). Puget Sound. Tropidonotus concinnus Hallow. (1852). Oregon. EuT^NiA pickeringi B. and G. (1853). Puget Sound. EuT^NiA LEPTOCEPHALA B. and G. (1853). Puget Sound. Eutcenia dorsalis B. and G. (1853). Rio Grande, Texas. EuT^NiA ELEGANS B. and G. (1853). El Dorado county, Cal. EuTJENiA VAGRANS B. and G. (1853). Cahfornia. Tropidonotus trivittatus Hallow. (1853). California. Eutcenia couchi Kenn. (1857). Pitt river, Shasta county, Cal. Eutcenia atrata Kenn. (1860). California. Eutcenia cooperi Kenn. (1860). Washington (?). EuT^NiA HAMMONDi Kcuu. (1860). San Diego county, Cal. E. sirtalis tctratcenia Cope (1875). Pitt river and Puget Sound. Eutcenia henshawi Yarrow (1884). Walla Walla, Wash. EUT.ENIA BiscuTATA Cope (1883). Klamath Lake, Oregon. E. elegans plutonia Cope (1892)=type of E. henshawi Yarr. E. elegans brunnea Cope (1892). Fort Bid well, Cal. E. elegans lineolata Cope (1892). No definite type. E. inferncdis inferncdis Cope (1892). Fresno and San Francisco. E. infernalis vidua Cope (1892)=types of E. atrata Kenn. E. sirtalis trilineata Cope (1892). Port Townsend, Ore. These names, twenty-three in number, appear to me to be reducible to three species, two of which present three forms each, sufficient in num- ber and constant enough in character, or so associated with a restricted area of distribution as to compel recognition as subspecies; or seven forms in all,^ the special features of which are capable of being arranged in a key which will cover all cases but the few anomahes upon which individual judgment is required. A. — Scales in 21 rows ; upper labials 8 : a. — Body moderately stout; head and eye rather small; posterior chin-shields equal or very little longer than anterior (E. elegans) : ^ Mr. Van Denburgh admits the same forms, but like values are not attached to them in all cases. Mr. Boulenger reduces them still further to five. As the only purpose of this paper is to analyze the various type specimens, no attempt is made to give references bej-ond tlie original descriptions. 288 PROCEEDINGS OF THE ACADEMY OF [March a' . — Usually 1 preocular: Black with three light stripes; or brown or red with stripes and spots distinct ; no nuchal blotches ; head smaller, E. e. elegans. Greenish-olive; stripes and spots not very distinct; nuchal blotches often present; ventrals often dark; head larger, E. e. vagrans. h'. — Usually 2 preoculars; often 23 rows; color hke vagrans or darker, E. e. hiscutata. b. — Body slender; posterior chin-shields much longer than an- terior; brown, usually without stripes; spots small and irregular, E. hammondi B. — Scales in 19 rows; upper labials 7; posterior chin-shields much longer than anterior {E. sirtalis) : a. — Body stout; head moderately large: Brown or black ; upper row of spots often fused into a stripe ; usually red on sides, E.s.parietalis. Usually black, with three light stripes, . . E.s. pickeringi. b. — Body small ; head and eye small ; often 17 rows and 6-8 labials ; color variable, E.s. leptocephala. Eutaenia elegans. Baird and Girard, Catalogue of North American Serpents, p. 34 (1853). This species has in nearly every case 8 upper labials and 21 rows of scales, of which the outer is either smooth or very faintly keeled, but an occasional example has 19 rows, or in one form 23, and the labials are sometimes 7. Lateral stripe on the second and third rows. The head is small and delicately formed ; the eye moderately small, and the poste- rior chin-shields are about equal to the anterior ones or very little longer. The throat and chin arc almost invariably yellowish. It is to be observed that two species found together in California, E. elegans and E. sirtalis parietalis, occa.sionally exhibit the scale and labial for- mula of the other, and very rarely the former may have posterior chin-shields as long as the shortest of the latter, and a very similar color variety occurs in each; but when elegans has 19 rows or 7 labials it may almost always be distinguished from any form of E. sirtalis l)y its short hinder chin-shields, and from any but E. s. leptocephala by its small head and eye. Three subspecies are to be admitted. Eutsenia elegans elegans. Eutania elegans B. and G., Cat. No. Am. Serp., p. .34 (18.53). Tropidonotus trivittntus Hallow., Proc. Acad. Nat. Sci. Phila., 1853, p. 237. E. elegans lineolata (part) Cope, Proc. U. 8. Nat. Mus., XIV, p. 6.55 (1892). E. infernalis infernalis Cope, /. c, p. 657 (not Coluber injernalis Blain.). Baird and Girard's type of E. elegans was almost black in color, with 1903.] NATURAL SCIENCES OF PHILADELPHIA. ''289 the three hght stripes well defined, and had 21 rows of scales; their second specimen had 19 rows and the dorsal spots were visible against the dark ground. Seven living garter snakes received at the Zoological Gardens of Philadelphia in May, 1902, from a collector at Oakland, Cal., throw much light upon the variations of this form. Specimens a. b, c, d, measuring from 270 to 290 mm. in length, coi respond in color to typical elegans, the three pale yellow stripes being on a brownish-black ground so dark that the spots arc barel}^ visible. Tvv^o have 21 dorsal rows and 8 upper labials; one has them 21 and 7; one has 19 and S. The three others are of mature size and much significance. Specimen e, 733 mm. long (tail 175), has 19 rows of scales, the outer weakly keeled, and 8 labials. Ventrals 163; subcaudals 87. The whole dorsal color between the spots, more or less of the outer row of scales, and the entire center of the belly are bright red. The spots are distinct and number about 94 in each row, to the vent. The dorsal stripe is orange-yellow and the laterals paler, more buff, but much marked with red. The small parietal spots and those on the ends of the ventrals, common in many species, are present and the upper por- tion of the labial sutures are shghtly margined with dark shading. The portion of the ventral surface not red is pale green. This specimen is the most brilliant and beautiful Eutcenia I have ever seen. Specimen /, 800 mm. long (tail 184), has 19 rows of scales, the outer faintly keeled, or smooth in places; 8 upper labials which are yellowish- olive slightly bordered with black. Ventrals 168 ; subcaudals 72. The dorsal stripe is bright yellow, and the laterals red. Body color dark brownish-black, the upper row of spots being entirely obscured, the lower row, 84 in number, showing as downward prolongations of the black from above, and separated by not very distinct patches of red just above the lateral stripe. This portion of the color pattern closely resembles some examples of E. s. parietalis. The belly is yellowish-olive in the center, slightly marked with red, the pattern of this part corresponding exactly to that colored red in e. The ends of the ventrals and the outer row of scales are olive-brown. No parietal spots and no distinct spots on the ends of the ventrals, but their bases are slightly dusky. This snake is E. infernalis infernalis Cope. Specimen g, 880 mm. long (tail 210), has 21 rows, the outer partly keeled and partly smooth, and 8 upper labials. Ventrals 167; sub- caudals 72. In pattern it is like e, but the body color is mostly brown instead of red, this last showing only on a single scale here and there, 19 290 ' PROCEEDINGS OF THE ACADEMY OF [March^. especially on the lateral stripe. There are about 90 spots in each row, the upper not very distinct. The belly is pale olive with small spots on the ends of many ventrals. No parietal spots and the labials very slightly margined. All these specimens have the small fine head, the short posterior chin-shields and the yellow throat and chin of E, elegans, and notwithstanding their wide difference in color, their specific unity is established by the fortunate fact that specimen g was a female,. gravid at the time of death, and I removed from her thirteen young, twelve of which are typical elegans in color, being so dark that the spots are barely visible, and corresponding exactly to specimens a. b, c, cL The thirteenth is also dark, but when first taken from the mother showed red markings on the flanks like those of specimen /. This red has almost entirely disappeared after twelve weeks in spirits. All of these young snakes have 21 rows of scales; ten have 8 labials; three have 7 on one side. The temporals range from 1-2-3 to 2-3-1. We have, then, unborn young exhibiting the colors of elegans B. and G. and infernalis infernalis Cope, the one resembling the latter having the scales and labials of elegans, contained in the oviducts of a female which, differing from both in color, departs in no other way from the features of elegans. If the red, which is a purely individual acquisition, were omitted, specimen g would come very close to E. e. lineolata Cope, some of which are referable to the present subspecies and some to E. e. vagrans. Trcpidonotus trivittatus Hallow, is no more than a typical elegan.f with 19 rows. E. e. elegans appears to occur only in California, from Shastha in the north down to the San Bernardino moimtains, beyond which it seems not to extend. Eutaenia elegans vagrans. Eutcenia vagrans B. and G., Cat. No. Am. Serp., p. 35 (1853).* Eutoenia couchi Kenn., U. S. Pac. R. R. Surv., X, Pt. 4, p. 10 (1857). Eutcenia henshawi Yarrow, Proc. U. S. Nat. Mus., VI, p. 152 (1884). E. elegans plutonia Cope, Proc. U. S. Nat. Mus., XIV, p. 653 (1892), and Rep. U. S. Nat. Mu.s., 1898, p. 1035. E. elegans hninnea Cope, Proc. U. S. Nat. Mus., XIV, p. 654 (1892). E. elegans lineolata (part) Cope, I. c, p. 655. This form ranges over the higher portion of the great plains, from whence it has penetrated through the valleys of the Snake and the Co- lumbia rivers to the western coast. It is the garter snake of the Sierras and the Rocky mountains. I have taken it at 6,000 feet eleva- ^ Eutwnia kennicotti Jan (Arch. Zool. Anat. and Phys., Ill, 1865, p. 216), with 21 rows and 8 labials, may belong here, but the description is vague, and the only locality given is northern United States. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 291 tion in Montana, and two of the Academy's specimens were collected by Dr. Henry Skinner at Sapello, New Mexico, at an altitude of 10,000 feet. The scutellation is similar to E. e. elegans, but the body is on an average larger and of stouter build. The head is larger and the eye relatively small. The color is almost always greenish or light olive, with the pale yel- low stripes narrow, not distinct, and often encroached upon by the small and ill-defined spots. An occasional example is brown, but vagrans, notwithstanding its extensive range, varies less than elegans. Usually there is a pair of dark nuchal blotches and the ventrals are more or less clouded with dark slate. While ordinary specimens are sufficiently distinct from elegans, there are intermediates in the western portion of its range, such as those called E. e. lineolata Cope, which reduce elegans and vagrans to the rank of subspecies. No examples of E. e. brunnea Cope are known but the two types collected at Fort Bidwell, Cal. These anomahes might almost as well be regarded as E. e. elegans, but their robust build and the localit}' whence they came incline me to refer them to the present form. The type of E. couchi Kenn. came from Pitt river, Cal., three hun- dred miles north of the known limit of E. hammondi, with which it has commonly been associated, and I agree with Van Denburgh in regard- ing the specimen as an anomalous vagrans. There does not seem to be any reason to consider Cope's tj'pe of E. &. plutonia, from Walla Walla, Washington, as anything more than a melanistic vagrans. It is to be observed that Yarrow had already given the name of E. henshawi to this specimen, at the same time describing Cope's second example, from western Arizona, as E. vagrans •plutonia. Eutaenia elegans biscutata. EutcBnia biscutata Cope, Proc. Acad. Nat. Set. Phila., 1883, p. 21. In western Oregon and Washington, and especially in the humid northwestern portion of the last state, E. e. vagrans is largely replaced by snakes generally similar but with a decided tendency toward melan- ism, and having usually two or three preoculars and sometimes 23 rows of scales. The types of E. biscutata were almost black and came from Klamath Lake, Ore., the most easterly locality from which it has been known, and where the rainfall is heavy. The association of these tendencies with a restricted area of distribution seems to require that it shall be recognized as a subspecies of E. elegans. 292 PROCEEDINGS OF THE ACADEMY OF [Marcll, Eutaenia sirtalis. Coluber sirtalis L., Syst. Nat., Ed. X, p. 222 (1758). The usual scutellation in this species is 19 rows and 7 upper labials, a formula which is very constant in the east, but in western forms 17 to 21 rows, or 6 to 8 labials sometimes occur. The lateral stripe is on the second and third rows. Compared with E. elegans it is larger and stouter, with a moderately large head. The posterior chin- shields are much longer than the anterior. East of the Mississippi river red is rarely developed, but I have seen one E.s. sirtalis hom'!>lorth Carolina which showed much of that color upon the sides, and another from Pennsylvania, in my own collection, is marked slightly with it on the flanks, but from the great plains westward there is a general disposition in reptiles to develop red, and it is more or less present in most subspecies of E. sirtalis from those parts. Eutaenia sirtalis parietalis- Coluber parietalis Say, Long's Exp. to Rocky Mts., I, p. 186 (1823). Coluber infernalis Blainville, Nouv. Ann. Hist. Nat., IV, 1835, p. 291, PL XXVI, fig. 3. Tropidonotus concinnus Hallow., Proc. Acad. Nat. Sci. Pliila., 1852, p. 182. Eutoenia dorsalis B. and G., Cat. No. Am. Serp., p. 31 (1853). Eutcenia ordinoides B. and G., Z. c, p. 33. E. sirtalis tetratania Cope (Fide Yarrow) ; LT. S. Geol. Surv. W. of 100th Mer., V, p. ,546 (1875), and Proc. U. S. Nat. Mas., XIV, p. 664. E. sirtalis trilineata (part) Cope, Proc. U. S. Nat. Mus., XIV, p. 665 (1892). E. s. parietalis has a range quite as wide as E. c. vagrans, for while it is absent from Arizona, and appears to be rare in Utah, it extends farther to the east on the plains. The scutellation is 19 rows and 7 labials, but now and then it has S labials, and more rarely 21 rows. It is ordinarily dark brown, with the spots more or less obscured and the lower row separated by red which shows on the skin and usuaUy on the scales in life, though much of it quickly fades in spirits. The small dark spots on the ends of the ventrals and the dark margins to the labials are either present or absent. The belly is yellow, various shades of slate, or almost black in dark specimens. The outer row of scales is often smooth, but sometimes weakly keeled. There is much difference in the amount and distribution of the red, especially in those from the Pacific coast, and there is a tendency toward melanism, strongly marked in the moist region of Oregon and Washington. In conformity with the law of color development in snakes this excess of pigment shows first upon the dorsal surface adjacent to the vertebral stripe, sometimes obscuring part or the whole of the upper row of spots (E. dorsalis) ; when it extends down far enough to reach the upper border of the lower row, these are left as 1903.] NATURAL SCIENCES OF PHILADELPHIA. 293 downward prolongations of the upper dark area, with red markings Ijetween {E. parietalis) ; when it extends to the lateral stripe all the spots are obliterated (E. pickeringi). A very few individuals show the bottom part of the lower row also obscured, leaving a series of red spots above {E. concinna) ; or the red has spread longitudinally, forming a red stripe between two black ones (E. s. tetraicBnia) . Of the above, the few specimens of E. dorsalis came from the Rio Grande valley, in New Mexico and Texas, except one fugitive noted by Cope from Portland, Ore., thirteen hundred miles away. Of the three tetratcenia which Cope is able to name, two, in the U. S. National .Museum, came from Pitt river, Cal., while the third. No. 6,085 in the Academy's coUecrion, from Puget Sound, originally entered as con- cinna, has the lower black stripe broken up into spots anteriorly. A smaller snake in the same jar as this specimen, and apparently collected with it, is an ordinary parietalis. Hallowell's type of Tropidonotus concin.nus (No. 6,324, Academy collection) is also marked on the label "tetratcenia'' by Cope. All these specimens have now been fifty years in spirits. The three concinna cited by Cope all came from western Oregon. With the exception of E. s. pickeringi, these selections of special cases in a physiological process appear to me quite arbitrary, and if the correctness of the method is once admitted, an indefinite number of others may as well be allowed. Baird and Girard's type of Tropidonotus ordinoides' came from Puget Sound, and the original description cannot be reconciled with tha't given of California specimens under the same name by these authors in their Catalogue of the following year. The type was prob- ably an E. s. leptocephala. Those subsequently described had 19-21 rows of scales, and the chief difference from ordinary parietalis was that the lateral spots were reddish-brown instead of red. A California example is figm-ed by Baird* with the form of head and the long posterior chin-shields of parictcdis, and 8 labials, but as the last-named species sometimes exhibits this number, it seems safe to refer ordinoides here, rather than to regard it with Cope as a subspecies of E. elegans. E. infernalis inferncdis Cope has been shown to be E. e. elegans, but Coluber infernalis Plain, occupies a somewhat doubtful position through the insufficiency of the original description and plate. Bocourt^ adds that it has 19 rows and 7 labials, which is the common formula for ' Proc. Acad. Nat. Sci. Phila., 1852, p. 176. 8 Pac. R. R. Report, PI. XXVI, fig. 3. '■> Bull. Soc. Zool. de France, 1892, p. 40. 294 PROCEEDINGS OF THE ACADEMY OF [March, parietalis, while his figures of Blainville's specimen^*' suggest this species rather than elegans; and as Cahfornia examples of both these forms sometimes exhibit the infernalis style of coloring, I see no reason at present to regard Blainville's type as anything more than the present species. Eutaenia sirtalis pickeringi. Eutamia 'pickeringi B. and G., Cat. No. Am. Serp., p. 27 (1853). E. sirtalis trilineata (part) Cope, Proc. U. S. Nat. Mus., XIV, p. 665 (1892). In the extreme west of Oregon and Washington, and especially about Puget Sound, a region of much moisture, melanism reaches an extreme and fairly constant degree in E. s. 'pickeringi B. and G. and E. s. tri- lineata Cope, the differences between which are trivial." E. s. pickeringi seems to be entitled to rank as a well-marked geo- graphical form, always so dark as to obscure the spots, stripes usually narrow, very distinct and variable in color, occasionally a little red on the sides, and the ventral surface always more or less dark, sometimes entirely 1^1 ack. Eutaenia sirtalis leptocephala. (t)Tropidonotus ordinoides B. and G., Proc. Acad. Nat. Sci.Phila.,1852, p. 176. Eidcenia leptocephala B. and G., Cat. No. Am. Serp., p. 29 (1S53). Entania atrata and E. cooperi Kenn., Pac. R. R. Surv., XII, Pt. 2, p. 296 (1860). E. infernalis vidua Cope, Proc. U. S. Nat. Mus., XIV, p. 658 (1892). This variable form is somewhat small in size and relatively stout; the head and eye are notably smaller than in other subspecies of E. sirtalis, and even more so than in E. e. elegans. The rows of scales are 17 or 19,theoutereithersmoothor keeled; upper labials usually 7, but sometimes 6 er 8; preoculars occasionally 2 and in a few cases 3. Posterior chin-shields much longer than the anterior. The color is olive, greenish or blackish-brown, the three light stripes variable in color and sometimes absent; spots small and hardly to be seen in dark specimens ; labials sometimes narrowly margined; parietal spots present; ventrals 139-152, yellow, greenish or dark slate. Some indi- viduals with 19 rows and 7 labials so nearly resemble certain phases of E. s. parietalis or E. s. pickeringi, and in fact some eastern E. s. sirtalis, that I cannot regard it as more than a subspecies, occupying British Columbia, western Washington and Oregon and northern California. E. cooperi Kenn. is clearly referable to this form. It has already been pointed out that the type of Tropidonotus ordinoides B. and G. ^'^ Miss. Sci. au Mexique, etc., PI. 55, figs. 2, 2a, 2b. " Cope's examples of trilineata from Fort Benton, Montana, are probably E. s. parietalis. 1903.1 NATURAL SCIENCES OF PHILADELPHIA. 295 (1852) was perhaps a leptocephala, but the fact cannot now he verified, and a mere possibihty should not be allowed to disturb existing nomenclature. ■ E. injernalis vidua Cope was founded upon Kennicott's origmal specimens of E. atrata. One of these is now No. 6,359 (original num- ber 970) in the collection of the Academy of Natural Sciences, and there is a second, No. 6,584, marked vidua in Cope's handwriting. Both of these specimens are labeled "San Francisco." I have elsewhere stated^2 that Cope's description is not accurate in details and have given my reasons for assigning these specimens to leptocephala. Mr. Van Denburgh considers them toheE.e. clegans, and states that this color form has been found only on the coast slope of the peninsula ()f San Francisco, and questions the occurrence of leptocephala in Cah- fornia. In consequence, I have reexamined the two examples of vidua and am still inclined to refer them to leptocephcda, leaving the geogra- phical part of the problem for further investigation.^^ In any event vidua would be no more than a synonym of atrata. Eutaenia hammondi. Eutwnia hammondi Kenn., Proc. Acad. Nat. Sci. Phila., 1860, p. 332. Hammond's garter snake does not range north of Fresno county, Cal., but extends southeastward into the plains of Arizona. The scutellation is that of E. elegans, but the body is slender, the head is narrow and elongated and the posterior chin-shields are much longer. The color is grayish or olive-brown; dorsal stripe narrow, indistinct or absent; the spots are always indistinct and^ sometimes absent, though indicated by black dots on many scales. Ventral sur- face yellowish, often with dark bases to the scuta, and at times clouded with slate toward the tail; this usually forms a line along the sutures between the subcaudals. Parietal. spots and nuchal l^lotches present; labials dark bordered, and a more or less evident pale post-oral crescent. This form has been regarded by some authors, including myself, as a subspecies of E. elegans, but further study of fresh material has satisfied me that it is distinct enough in character and geographical range to be admitted to specific rank^ __^__ 12 Proc. Acad. Nat. Sci. Phila., 1901, p. 30. . '■' There may be a doubt as to the exact localities where the specimens on which vidua was founded were collected. Kennicott g;ave none in the original descrip- tion of E. atrata, but in a footnote Dr. Cooper, the naturalist of the survey, says "California " The one in the Academy's collection is labeled ban l^rancisco, but there is also in the collection an undoubted leptocephala of about the same date, bearing a like label. San Francisco has always been the point at which all California interests center. 296 PROCEEDINGS OF THE ACADEMY OF [Maxch, The only other name with which it has been confused is E. couchi Kenn., which has been shown to be in all probability an aberrant E.e. vagrans. The geographical relations of the species here discussed are as fol- lows : E. e. elegans meets with E. e. vagrans and E. s. parieialis through- out its range. At its southern extension it is overlapped by E. ham- niondi and at its northern it touches E. s. Icptocephala. E. e. vagrans meets E. hamnwndi in California, and is in contact with E. s. parie- talis almost everywhere except in Arizona, although the latter does not appear to ascend as high in the mountains. In the northwest it merges into E. e. hiscutata and reaches E. s. leptocephala. On the plains it meets E. radix, and in the southwest comes more or less in touch with E. marciana, E. eques and perhaps E. 7negalops. E. s. parietalis meets all of these except the last. In the east it overlaps scattering examples of E. s. sirtalis and probably E. proxima. In the far northwest it merges into E. s. leptocephala and E. s. pick- eringi. E. hammondi meets with E. marciana in the eastern part of its range in Arizona. The connection between moisture and variability, especially in the direction of color intensity, may be profitably observed in these snakes. More than a million square miles of the territory occupied by the widely ranging E. e. vagrans and E. s. parietalis lies east of the high mountain chains of the Pacific coast and has an annual rainfall of from 10 to 25 inches. Over this vast expanse the characters of these two species are very constant, and variation has contributed to the list of synonyms only E. dorsalis B. and G., in which melanism is in its earliest stage, and the one specimen of E. henshawi Yarr. (=E. e. plutonia Cope) from Walla Walla. The region of great moisture, with a rainfall of from 50 to over 100 inches, occupies not more than a hundred thousand square miles, extending from latitude 40° in northern California to British Columbia. The type localities of the following five forms, characterized by pro- nounced melanism and often an excess of red, all fall within this re- stricted area: E. ordinoides, E. concinna, E. pickeringi, E. s. tetrcdetnia and E. s. trilineata. In addition to these, E. e. hiscutata and E. s. leptocephala, found there, also show marked tendencies to develop dark colors as well as instability in scutellation. Five forms — E. infernalis Blain., E. i. inferncdis Cope, E. i. vidua, E. atrata and E. e. elegans — were described from the neighborhood of San Francisco, where the actual rainfall does not exceed 25 inches, 1903.] NATURAL SCIENCES OF PHILADELPHIA. 297 but where much moisture is carried over the coast belt and up the river valleys by the persistent fogs which sweep in from the Pacific ocean. The region occupied by E. hammondi is exceedingly dry, with a rainfall of less than 10 inches, and tliis species is relatively constant and is not known to show any tendency to melanism.^* '* That humidity in some way influences the metabolic processes which lead to pigmentation can hardly be doubted. Temperature need scarcely be con- sidered in the present case, for the dry region, extending from Arizona to nortliern Montana, and to considerable elevations, has a very great thermal range, while the wet region is relatively equable. There is a suggested connection between the large amount of uric acid produced by reptiles and the fact that the yellow and orange coloring matter from the wings of certain butterflies has yielded a substance closely related to uric acid, but physiological chemistry is not yet competent to explain how these waste products are converted into pigments. The liberty to indulge in the striking colors developed in the garter snakes of this region is partly due to the protection afforded by luxuriant vegetation, and perhaps in some degree to the absence of the three snake-eating genera Spiloies, Ophibohcs and Elaps. 298 PROCEEDixGS OF THE ACADEMY OF [March, A NEW SPECIES OF NYCTALEMON. BY HENRY SKINNER. Nyctalemon curvata -f n. sp- Plate XIII. Upper side. Primaries dark smoky-brown, almost Ijlack. The costa is crossed by a number of very narrow, pale slate-colored lines which break up the ground color into spots. A distinct white band or fascia crosses the wing from the costa to the interior margin; this band com- mences on the costa 30 mm. from the base of the wing, measured along the costa; it ends on the interior margin 34 mm. from the base of the wing. The band is 3 mm. wide at the costa and gradually grows narrower. Its white color ends before it reaches the exterior margin. The iDand is distinctly divided into seven white spots by the dark nervures and it is markedly curved, the concave side being in- ward. There are no other markings except that the apices are slightly lighter in color. Secondaries. The wings are the same color as the primaries and are crossed by a band or fascia some shades lighter than the ground color of the wing; from this toward the base are no markings. Beyond the band toward the outer margin the wings are lighter in color and are marked with a long l^lack shade and a number of somewhat linear black markings. The exterior margin is l^roken into two small and one long point and there is also a long broad tail to the wing. Commencing in the larger point is a black band about 2 mm. wide and forming the edge of the wing and extending to the in- terior margin. It crosses the larger tail al^out 10 nnn. from the ti]"). leaving the latter white from where it crosses to the end. Under side. Primaries. The ground color of the wings is much lighter; the white band is the same as above, except that it extends unbrokenly white to the exterior margin. There is a black shade beyond the ])and on the apex side and numerous black, somewhat linear markings ; the tip of the wing is whitish. There is also a black shade on the inner side of the band and the linear markings are also found toward the l)ase of the wing. The secondaries are marked in the same general way and the l^and is distinctly white l)ut narrower than on the primaries. It terminates in an irregular black band-like spot which reaches the interior margin. Expanse (from body to apex of wing on one side) 62 mm. The very dark color, the comparatively wide and very distinct, curved white band serve to distinguish this species from the others in the genus. DescrilDed from two speci- mens presented to the Acadeni}^ of Natural Sciences of Philadelphia bv Prof. Levi W. Mengel. They are from Tonga, off Santo, New Hebrides, Occanica. 1903.1 NATURAL SCIENCES OF PHILADELPHIA. 299 STUDIES IN AMERICAN FOIIFICULID.5:. BY JAMES A. G. REHN. The material treated in the following pages is contained in the collection of the United States National Aluseiun. The privilege of examining this series was secm'ed through the kindness of Mr. William ■ H. Ashmead,, Assistant Curator of Insects in that institution. Family FORFICULID^. The general arrangement of the species under this family is accord- ing to the system used by De Bormans and Krauss.^ The scheme of classification recently proposed by Verhoeff^ has not been considered, as the WTiter has made no investigations into the characters there brought into prominence. Genus PYRAGRA Serville. 1S31. Pyragra Serville, Ann. Sci. Nat., XXII, p. 34, Type: P. fuscata Serville. Pyragra fuscata Serville. 1831. Pyragra fuscata Serville, Ann. Sci. Nat., XXII, p. 34. [Cayenne.] One female; San Carlos, Costa Rica. (Schild and Burgdorf.) [U. S. N. M.] The use of Dohrn's later name saussurci for this species appears un- warranted, as Serville's specific name is based on a description which is fairly complete, and perfectly apphcable to this form. This species has been recorded from localities between south central Mexico (Cordova and Atoyac) and French Guiana. Geuus ECHINOPSALIS De Bormans. " 1893. Echinopsalis De Bormans, Biol. Cent.-Amer., Orth., I, p. 2. Type: E. guttata De Bormans. 1 Das Tierreich, 11 Lieferung: Forficulida? und Hemimeridse, Berlin, 1900. 2 Zoologischer Anzeiger, XXV, pp. 181-208; and Sitzungs-Bericht der Gesellschaft naivrforschender Freunde, Berlin, 1902, pp. 7-18. 300 PROCEEDINGS OF THE ACADEMY OF []\Iarch, Echinopsalis brevibractea n. sp. Type: cJ^(immature) ;Motzorongo, Vera Cruz, ]Mexico. January, 1S92. (H. Osborn.) [U. S. N. M.] This form apparently belongs to the genus Echinopsalis, though not wholly agreeing with the generic diagnosis. Difference from the generic characters is noticed in the longer third antenna! joint, the subtruneate posterior margin of the pronotum, the shorter anal seg- ment and in the longer, more attenuate forceps. No close relationship exists with the only previously known species, E. guttata De Bor- mans from Chontales, Nicaragua. Body of moderate length, considerably depressed; surface subscab- rous, covered with short, stiff, mostly erect hairs. Head rather flat, slightly narrower than posterior ^^ddth of tlie pronotmn, the posterior margin subemarginate centrally; basal joint of the antennae inverted conoid, shorter than the uniform cylindrical third joint, fourth and succeeding joints (beyond eleventh missing) short and thick, bead-like, broader than long. Pronotum subquadrate; anterior margin broadly rounded, the angles not at all apparent; posterior margin subtruneate; median sulcus well marked and extending the entire length of the pronotum; lateral margins considerably raised. Tegmina and -^dngs only partially developed, but in such condition the former almost equal the pronotum in length. Anal segment moderately deep, over twice as wide as long, very heavily hirsute. Subgenital plate broad, the apex gently rounded. Forceps trigonal in section; rather elongate, parallel and unseparatecl; the inner margins finely toothed, this being stronger and more apparent basally; apical portion moderately incurved, crossed. Femora rather robust, uncurved; tibiae moder- ately slender, not equalling the femora in length. Tarsi moderately compressed, the first and third joints about equal in length. General color wood-brown, the limbs being mottled with touches of dirty-yellowish. Measurements. Total length (approximately), 9.5 mm. Length of pronotum, 2.5 " Width of pronotiun, 2.5 " Length of anal segment, 1.5 " Length of forceps,] 3.5 " Four specimens, all from the type locality, have been examined. In one of these eighteen joints are present in the antennae, twentj'-two being credited to E. guttata; this possibly being true of hrevibractca in the unbroken antennae. 1903.] ' NATURAL SCIENCES OF PHILADELPHIA. 301 Genus LABIDUEA Leach. "1815. Lahidura Leach, Edinburgh Encyc, IX, p. 118." Labidura bidens (Olivier). 1791. Forficula bidem Ohvier, Encyc. Meth., Hist. Nat., VI, Pt. II, p. 466. [Jamaica.] Twelve specimens; six males, two females, four specimens damaged and sex uncertain. Bermuda. [U. S. N. M.] (damaged). Cabanas, Cuba. June 2. (Palmer and Riley.) [U. S. N. M.] Guanajay, Cuba. April 30. . (Palmer and Riley.) [U. S. N. M.] ■ This name has recently^ been revived for the form occurring in the West Indies, Mexico and Colombia. De Bormans and Krauss place it as a synonym of pallipes Fabricius, but as Kirby says, the latter name ''is usually apphed to the dark form of L. riparia; the types are lost, and the insect cannot be identified till more specimens are received from its locahty, the Cape Yerde Islands." The specimens from Bermuda are badly damaged, and are placed under this name with some uncertainty. Genus PSALIS Serville. 1831. Psalis Serville, Ann. Sci. Xat., XXII, p. 34. Included P. americana (Palisot) and P. morhida Serville. Psalis americana (Palisot). 1817. For-ficula americana Palisot, Ins. Rec. en Afr. et Amer., p. 165, Orth., t. 14, fig. 1. [San Domingo.] Ten specimens; nine males, one female; San Carlos, Costa Rica. (Schild and Burgdorf.) [U. S. N. :\I.] This large and striking species has been recorded from locahties extending from Chontales. Nicaragua, to western Colombia and Vene- zuela, and also from Cuba and Haiti. In the mature specimens (6) in this series a shght variation is noticed in the internal margins of the forceps. In four specimens these are slightlv extended basally and distinctly crenulate. In two these parts are straight and with the crenulations hardly visible. The im- mature speciiuens (all males) are very interesting and might easily be mistaken for members of the genus Anisolabis. Two stages, rather different in general appearance, are represented in the series and are worthy of separate description. One conclusion reached from a study of these specimens is, that the immature forms possess lateral folds on the abdominal segments which practically disappear by the time the ad ult stage is re ached. 3 Ann. and Mag. Nat. Hist., 7th ser., XI, p. 66, January, 1903. ^02 PROCEEDINGS OF THE ACADEMY OF [Marchj Size medium (25 mm.); body fusiform; surface rather polished, the abdomen and forceps finely punctate, the whole sparsely beset with stiff bristles, and the abdominal segments fringed with fine hairs. Head considerably broader than the pronotum. Pronotum slightly longer than broad; the anterior margin subtruncate, the posterior rotundate; the median region bearing a shallow sulcation, deepest an- teriorly. Meso- and metanotum with the tegmina and wings incom- pletely developed. Abdomen rather broad, the four anterior segments with the lateral folds but slightly developed. Anal segments large, subquadrate, the posterior constriction being slight; median sulcus slightly developed posteriorly. Forceps of medium length, robust, triangular in section, the right more curved and less hooked than the left, and crossing the latter; internal margins minutely undulate, ex- hibiting no signs of true teeth. Femora short and robust, the greatest thickness being toward the base. Tibiae slightly shorter than the femora in the median pair of limbs, equal in the others. General color blackish-brown; clypeus, labrum, mandibles, tibige and tarsal joints dirty yellow (gamboge); femora of the general color with an apical ring of dirty yellow. Antenna? dull-brown with the first, second, thirteenth and fourteenth joints yellow. Measurements. Largest Average of three Specimen. specimens. Total length, 25.0 mm. 22.7 mm. Length of pronotum, 3.2'' 28" Width of pronotum, 3.0 " 26 " Length of forceps, 45 " 37 u Length of anal segment, 3.1 " 2 7" Anterior width of anal segment, ... 5.1 " 4.2 " The other stage, which represents a lower and more undeveloped type than that described above, is represented in the series by a single specimen. Size rather small (13.5 mm.); body subfusiform, the greatest width being abdominal; surface moderately polished and with stiff bristles generally distributed, the apical dorsal segments of the abdomen being margined with fine hairs, which are also present on all the ventrals. Head rather elongate, broader than the pronotum. Pronotum about as broad as long, the anterior margin subtruncate, the posterior ro- tundate; median sulcus distinctly marked. Mesonotum posteriorly truncate; metanotum centrally emarginate. Abdomen broad, each dorsal segment (except the apical three) with the posterior n'largin 1903.] NATURAL SCIENCES OF PHILADELPHIA. 303 bearing a series of folds or plaits, which are generally disposed in rows of four on each side, these being most numerous anteriorly and vanishing posteriorly. Ventral surface of the abdomen wth the segments apically with small nodes, somewhat like the folds of the dorsal surface, but these being very irregularly distributed. Such folds and nodes are, in many cases, set with stiff bristles. Anal segment not very long, rather broad, the surface almost flat; median sulcus very broad and shallow. Forceps of medium length, rather robust, triangular in section, the left exceeding the right in length; internal margins not toothed, almost straight. Femora short and robust; tibiae rather compressed, heavy, shorter than the femora in the posterior and median pairs, equal in the anterior pair. General color blackish-brown; labrum, portions of clypeus and man- dibles, tibise and tarsal joints dirty yellow (gamboge); femora of the general color becoming suffused with yellowish apically; antennae wood-brown, the apical portion of basal joint washed with dull- yellowish. Measurements. Total length, 13.5 mm. Length of pronotum, 2.0 " Width of pronotum, 2.0 " Length of forceps, 3.0 " Length of anal segment, 1.5 " Anterior width of anal segment, 2.7 " Psalis pulchra n. .^p. Type:c?; San Carlos, Costa Rica. (Schild and Burgdorf.) [L'. S. N. M.] This species is apparently closest allied to P. americana and gaga- thina, both of which very considerably surpass it in size, and from which it differs in the more elongate pronotum, the comparatively heavier forceps and different coloration. Relationship appears to exist with P. rosenhergi Burr from Ecuador, but pulchra may readily be separated by color as well as the non-pilose body. Body of medium size (for the genus) ; elongate fusiform ; surface moderately polished. Head broader than the pronotum; antennae 15 to 16-jointed, third joint shorter than the basal, fourth and fifth rather small and rounded, sixth to eighth elongate, increasing in size, ninth to extremity equal to the eighth in size and shape — the apical portion of each segment being enlarged and of considerable greater size than the base. Pronotum longer than broad and slightly con- stricted posteriorly; anterior margin subtruncate, the posterior roundly produced centrally, lateral margins considerably deflected upward ; 304 PROCEEDINGS OF THE ACADEMY OF [March, median sulcus distinctly marked except at the extreme posterior mar- gin, which is strongly depressed and flattened, a condition also found in the lateral margins. Tegmina of considerable length, reaching to the base of the posterior femora; the anterior shoulders broadly rounded; extremities subtruncate. Exposed portions of the wings about half as long as the tegmina, the posterior and external margins broadly rounded. Al^domen with the lateral margins almost parallel, the basal segments bearing on their margins minute nodes, each of which gives rise to one or more bristles. Anal segment deep, subrec- tangulate in outhne; median line hardly visible; the apical section bearing, between the bases of the branches of the forceps, a transverse depression. Subgenital plate triangular in outline; apex broadly emarginate, and provided with long, soft hairs. Forceps short and heavy, not quite equal to half the exposed abdomen in length, tri- angular in section l^asally ; apical portion strongly incurved ; the right arm longer than, and overlapping, the left, internal right edge finely crenulate, a minute tooth being developed just before the base. Femora moderately heavy, compressed; tibiae about equal to the femora in length; third tarsal joint shorter than first, the first and second joints heavily haired below, second joint very small. General color deep mahogany-brown, becoming dirty yellow on the limbs and lower surface of the head and thorax. Pronotum margined laterally with very duU-yellowdsh. Head reddish-brown. Antennse with the two basal joints reddish-brown; the tenth, eleventh and part of the twelfth pale-yellowish; remainder of the general tint. Tegmina bearing shoulder spots of sti-aw-yellow. Exposed wings with the basal half straw-yellow. Measurements. Total length, 19.5 mm. Length of pronotum, 2.5 " Width of pronotum, 2.0 " Length of tegmina, 4.5 " Anterior width of anal segment, 3.5 " Length of anal segment, 2.2 " Length of forceps, . . 3.0 " Genus ANISOLABIS Fieber. 1853. Anisolabis Fieber, Lotos, III, p. 257. Included A . tnaritima and mcesta. Anisolabis annulipes (H. Lucas). 1847. Forjiccsila annulipes H. Lucas, Ann. Soc. Ent. France, 2e ser., V, Bull., p. Ixxxiv. [Jardin des Plantes, Paris ; supposed to have been trans- ported from North America.] One male; Minatitlan, Vera Cruz, Mexico. February 1, 1892. (H. Osborn.) [U. S. N. M.] 1903.] NATURAL SCIENX'ES OF PHILADELPHIA. 305 Anisolabis antoni (Dohrn). 1864. Forcinella Antoyii Dohrn, Entom. Zeitung, Stettin, XXV, p. 289. [Venezuela.] Two females : Mexico. (Duges.) [U. S. N. M.] Minatitlan. Vera Cruz, Mexico. February 1, 1892. (H. Osborn.) [U. S. N. M.] These specimens measure 15 and 12.5 millimeters in total length. This species has been recorded from San Luis Potosi, Mexico; Cape- tillo, Guatemala ; Chontales, Nicaragua and Venezuela. Genus LABIA Leach. "1815. Labia Leach, Edinburgh Encyc, IX, p. 118." Labia guttata Scudder. 1876. Labia guttata Scudder. Proc. Boston Soc. Nat. Hist., XVIII. p. 265. [Texas.] One female; Orizaba, Vera Cruz. Mexico. January 9-16, 1892. (H. Osborn.) [U. S. N. M.] This specimen does not wholly agree with Scudder's description, but still seems referable to this species. Labia fiaviscuta n. sp. Type: ? ; Orizaba, Vera Cruz, Mexico. January 9-16, 1892. (H. Osborn.) [U. S. N. M.] Apparently allied to L. arcuala Scudder, but differing in the elongate pygidium, the shorter exposed portion of the wings, and the predomi- nance of yellow instead of black on the pronotum. From L. championi De Bormans, the only other form to which it appears at all allied, it can be immediately separated by the thirteen-jointed antennae, the shorter exposed portion of the wings, and the narrower pygidium. Body of very small size (5.5 mm.) ; greatest width at the apical por- tion of the abdomen; surface of the head and pronotum moderately polished, abdomen and tegmina dull. Antennae thirteen-jointed, the fourth and fifth joints much shorter and more bead-like than the apical segments. Pronotum slightly longer than broad ; anterior margin very broadly rounded; posterior margin gently rounded; laterals parallel; median sulcus reduced to a slight groove on the posterior half. Teg- mina short, not more than one and a half times as long as the pronotum, but slightly extended laterally, the shoulder angles well rounded; apex subtruncate. Exposed portion of the wng very small, apparent only as the extreme tip, and placed next the median suture. Abdomen with well-developed side folds present on two of the median segments. Anal segment and subgenital plate small, transverse, the apical margins 20 306 PROCEEDINGS OF THE ACADEMY OF [March, truncate. Pygidium slightly longer than basal breadth, diminishing in width, apex truncate. Forceps simple, unarmed, about equal to half the length of the abdomen; basal halves of the arms subparallel, apical halves bent at an angle and meeting, together forming an acute angle. General color brownish-black above, reddish-brown below. Pro- notum wdth the anterior two-thirds reddish-orange, posterior third brownish-black. Antennae wood-brown, with the two apical joints pale-yello^\ash. Limbs brownish-orange. Measurements. Total length, 5.5 mm. Length of pronotum, LO " Width of pronotum, LO " Length of tegmina, L2 " Length of forceps, LO " Labia rotundata Scudder. 1876. Labia rotundata Scudder, Proc. Boston Soc. Nat. Hist., XVIII, p. 263. [Mexico.] One female; Orizaba, Vera Cruz, Mexico. January 9-16, 1892. (H. Osborn.) [U. S. N. M.] This specimen exhibits one structural difference from the original description, in that the posterior angles of the pronotum are rather narrowly rounded, instead of "very broadly rounded." This species has been recorded from Mexico, Guatemala City, Acey- tuno and Duenas, Guatemala. Genus SPARATTA Serville. 1839. Sparatta Serville, Orthopt^res, p. 51. Type: S. peJvimetra Serville. Sparatta flavipennula n. sp. Types: c^ and 9 ; Motzorongo, Vera Cruz, Mexico. January, 1892. (H. Osborn.) [U. S. N. M.] Apparently allied to S. pelvimetra Serville, but differing in the broader and shorter pygidium, the shorter anal segment, the unarmed and sub- lamellate inner edge of the forceps, and different general coloration. Relationship also appears to exist with S. pygidiata and semiruja Kirby, but from the former it can be separated by the deeper anal segment, the longer pygidium and strongly toothed forceps ; and from semiruja by the same characters and the longer points to the forceps. d^ (not quite mature). Size small; body strongly depressed, broad- est in the abdominal region, sparsely clothed with rather long bristles. Head equal to the pronotum in width ; antennae with the first and third 1903.] NATURAL SCIENCES OF PHILADELPHIA. 307 joint subequal, fourth joint subglobose, sixth and succeeding segments (present to eighth) cyhndrical, the fifth joint representing a transition between the two types. Pronotura sHghtly longer than broad; ante- rior section produced centrally, the anterolateral angles being obtusely rounded; posterior margin subtruncate; greatest width of the pro- notum across the anterior angles; median sulcus very slight Anal segment transverse, the depth not exceeding one-third the width" posterior margin truncate; median section with a longitudinal sulca- tion. Pygidium triangular, subconoid, the apex truncate. Forceps about equal to the abdomen in length, parallel ; apical regions incurved, meeting, but not lapping; inner margins with a very slight lamellate edge on the basal two-thirds, this edge being crenulate and not dentate feubgenital plate transverse, slightly emarginate centrally. Femora robust, very heavily developed. Tibia shorter than the femora bowed. ' General color of the upper surface blackish-brown, slightly pohshed" head, antennae, forceps and limbs pale reddish-brown, deepest in the color on the Umbs. Under surface reddish-brown ; the subgenital plate being of the same pale tint as the forceps. 9 . Size medium; body strongly depressed. Head about equal to the pronotum in width; antennae imperfect. Pronotum longer than broad; anterior portion produced centrally, the antero-lateral angles obtusely rounded; posterior margin and postero-lateral angles broadly rounded; greatest width of pronotum between postero-lateral angles • median sulcus slight, obsolete posteriorly. Tegmina half as long again as the i^ronotum ; lateral margins parallel, the shoulders rounded ; apical portion, taken as a unit, rounded. Exposed portions of wings about half as long as the tegmina, rounded. Anal segment tra^'ns verse half as long again at anterior margin as deep, contracted, apical portion considerably narrower than basal. Pygidium half as long again a,s basal width, tapering, apex truncate. Forceps considerably less than the abdomen in length, parallel, apical portion moderately hooked- in- ternal margins being a slight lamellate edge, this edge being present on all but the apical fourth, though subobsolete basally ; basal internal sec- tion bearing a very distinct tooth. Subgenital plate deep, about half as broad again as long; apical margin very broadly rounded. General color of the upper surface blackish-brown, slightly polished • antenna, limbs, anal segment and forceps reddish-brown, deepest in color on the anal segment and forceps; exposed portions of the wings pale-yellowish, lined along the median edges with brown. Under surface reddish-brown. 1.5 1.5 1.1 2.5 308 PROCEEDINGS OF THE ACADEMY OF [March, Measurements. (^ ? Total length, 7.5 mm. l?-0 "^f^" Length of pronotmn, ^'^ ]] Width of pronotmii, ^'^ u Length of anal segment, 4 "^ Length of forceps, 1-5 An additional immature specimen from the type locahty was also examined. Genus ANCISTROGASTER Stal. 1855. Ancistrogaster Stal, Ofv. Kongl.-Yet. Akad. Forh., XII, p. 349. Type: A. luctuosus Stal. Ancistrogaster tolteca (Scudder). 1876. Forficula tolteca Scudder, Proc. Boston Soc. Nat. Hist., XVIII, p. 261 [Mexico.] Three males; Orizaba, Vera Cruz, Mexico. January 9-16, 1892. (H. Osborn.) [U. S. N. M.] Previous records: Orizaba and Atoyac, Mexico, and Coban, Guate- mala. Ancistrogaster gulosa Scudder. 1876. Ancistrogaster gulosa Scudder, Proc. Boston Soc. Nat. Hist., XVIII, p. 259. [Puebla, Mexico.] Fifteen specimens; eight males, seven females; Orizaba, Vera Cruz, Mexico. January 9-16, 1892. (H. Osborn.) [U. S. N. M.] Genus OPISTHOCOSMIA Dohru. 1865. Opisthocosmia Dohrn, Entom. Zeituug, Stettin, XXVI, p. 76. Type: By selection, 0. eenturio Dohrn. Opisthocosmia (Sarcinatrix) anomalia n. subg. et sp. Types: d^ and 9 ; Turrialba and San Carlos, Costa Rica. (Schild and Burgdorf.) [U. S. N. M.] This form appears to differ from all the other members of the genus Opisthocosmia, in the forceps of the male being parallel throughout their entire length. This, as well as the peculiar structure of the subgenital plate, has caused me to create a new subgenus for the reception of the species. (^ Size medium ; form rather slender and graceful ; surface polished and furnished with long hairs. Head rather broader than the prono- tum; interorbital region tumid and bearing two median longitudinal impressions; eyes prominent, ovate in outline; antenna? with the basal joint elongate, robust, of greatest thickness in the.distal portion, second 1903.] NATURAL SCIENCES OF PHILADELPHIA. 309 joint very small, less than one-fifth the length of the first third and ourth joints of eqnal length which is about one-third that of basal ioin b yond the fourth joint the segments are subcyhndncal gradually le'^ing in length, the total number being twelve. Pr— subequal anterior margin truncate, lateral margins subparaUel, poSo broadly rounded; fine of depression crossing the median fine posterior to the middle; median line shallowly sulcate m the anterior haU ithtly carinate in the posterior half. Tegmma broad, shoulder prLne'nt, narrowly rounded; lateral margins subparalle , apical po - Z very slightly arcuate. Exposed portion o wings - her dongat over halt as long as the tegmina, apex narrowly rounded Abdomen up rin.-; basal segments with the lateral folds well ^;^^^' internal borders strongly crenulo-dentate ; surface supplied with long h"remora and m. of subequal length, the anterior femora considerably heavier than the median and posterior pair. Tars, with the third ioint heart-shaped, flattened. % neral' color wood-brown; the median portion o the pronotiun the marc-ins of the tegmina and exposed portions of the wings, the aMomen forceps and antennal markings of a darker tint than the r"aMe'r of the body which is of a sienna tone. Antenna with the CTeld, tenth, eleventh and twelfth dark, the eighth partially so, S; the ninth has the basal portion dark, the remainder being ""ttedium, larger than the male; form rather robust. Abdo- men expanded the greatest width being median. Anal segment ILtr S bgenitll plate produced subtriangular. Forceps prac- tLX tte same fn structure as those of the male, but more elongate "Ceratcolor as in the male. Antennae with the basal and apical portions of the first, the second, seventh and eighth joints dark. Measurements. „ 8.5 mm. 10.5 mm. Total length, 11" 1.5 " Length of pronotum, 10" 1.2 " Width of pronotum, 90 " 2.7 " Length of tegmina, 15" 2.0 " Length of forceps, One additional female from Turrialba was also exammed. 310 PROCEEDINGS OF THE ACADEMY OF [March, Genus NEOLOBOPHORA Scudder. 1875. Neolobophora Scudder, Proc. Boston Soc. Nat. Hist., XVII, p. 281. Type: N. hogotensis Scudder. Neolobophora ruficeps (Burmeister). 1838. F[orficula] ruficeps Burmeister, Handb. der Entom., II, p. 7.5.5. [Mex- ico.] Two specimens; male and female; Orizaba, Vera Cruz, Mexico. January 9-16, 1892. (H. Osborn.) [U. S. N. M.] The forceps of the female are slenderer than those figured by De Bormans.^ Genus APTERYGIDA Westwood. 1840. Apterijgida Westwood, Introd. Class. Ins. Synop. Gen., p. 44. Type: Forficula pedestris Bon.=albipennis Meg. Apterygida linearis (Eschscholtz). 1822. Forficula linearis Eschscholtz, Entomographien, I, p. 81. [Sta. Cath- arina, Brazil.] Twenty-three specimens; fifteen males, eight females; Orizaba, Vera Cruz, Mexico. January 9-16, 1892. (H. Osborn.) [U. S. N. M.] El Guama, Cuba, March 6. (Palmer and Riley.) [U. S. N. M.] Bonito, Province of Pernambuco, Brazil. (A. Koebele, on cotton.) [U. S. N. M.] Turrialba, Piedras Negras and San Carlos, Costa Rica. (Schikl and Burgdorf.) [U. S. N. M.] This series of males exhibit considerable variation in the structure of the forceps, in some cases these appendages being constricted centrally. The representatives of this sex range in size from 15 to 20 mm. in total length ; in the females from 10 to 15 mm. Apterygida californica (Dohrn). 1865. F[orfi.cula] Californica Dohrn, Entom. Zeitung, Stettin, XXVI, p. 85. [California.] Two males : Orizaba, Vera Cruz, Mexico. January 9-16, 1892. (H. Osborn.) [U. S. N. M.] Bonito, Province of Pernambuco, Brazil. (A. Koebele, on cotton.) IJU. S. N. M.] While tentatively allowing this form specific rank, I am not at all satisfied that it is more than a mutation of linearis, specimens exam- ined being almost intermediate, having the internal tooth reduced to a minimum. * Biol. Cent.-Amer., Orth., I, PI. 2, fig. 8. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 311 The specimens above listed are rather smaller than any males in the series of linearis, being 14 and 11 mm. in total length. The previous records for this species are from California, northern Sonora and Teapa, Tabasco, Mexico. Genus FORFICULA Linnaeus. 1758. Forficula Linnseus, Syst. Nat., X ed., p. 423. Type: F. auricularia Linnaeus. Forficula lugubria Dohm. 1862. Forficula lugubris Dohrn, Entom. Zeitung, Stettin, XXIV, p. 230 [Cordova, Mexico.] Eleven specimens; four males, seven females; Orizaba, Vera Cruz, Mexico. (H. Osborn.) [U. S. N. M.] These specimens differ from the original description in their more reddish coloration. The previous records for this species cover from Durango, Mexico, to Valladolid, Yucatan, and Vera Paz, Guatemala. Forficula metrica n. sp. Type:d^; Orizaba, Vera Cruz, Mexico. January 9-16, 1892. (H. Osborn.) [U. S. N. M.] Closely aUied to F. lugubris Dohrn, but differing in the longer forceps, which also bear an elongate low ridge on their internal upper surface, instead of a comparatively high rounded projection as in lugubris. Size medium; form rather robust; surface glabrous, sparsely sup- plied with short hairs. Head about equal to the pronotum in width; interspace between the eyes rather tumid; antennae with but six joints present, the basal one moderately robust, the third to sixth of simple form, slightly increasing in length distally. Pronotum slightly broader than long; anterior border truncate, lateral borders slightly arcu- ate, the posterior subtruncate, the postero-lateral angles very broadly rounded ; line of depression but slightly marked and crossing about the middle; median hne shallowly sulcate anteriorly, narrowly carinate posteriorly. Elytra of moderate width; shoulder well rounded and not at all prominent; lateral margins parallel; posterior margin sub- truncate. Exposed portion of wings half as long as the tegmina, rounded. Abdomen with the median section considerably expanded; two median segments with the lateral folds developed, the posterior the heavier. Anal segment transverse, the posterior margin subarcu- ate. Subgenital plate broadly triangular, the apex truncate. Forceps almost equal to the remainder of the body in length ; the basal fourth parallel, depressed and bearing a lamellate ridge along the internal 312 PROCEEDINGS OF THE ACADEMY OF [MaTch, margin; apical three-fourths arcuate, unarmed. Pygidium very short, subacuminate. Median and posterior limbs with the femora and tibiae of subequal length; anterior limbs with the tibiae distinctly exceeding the femora. General color dark-brown; antennae and limbs gamboge-yellow. Measurements. Total length, 17.0 mm. Length of pronotum, 1.7 " Width of pronotum 2.0 " Length of tegmina, 3.0 " Length of anal segment 1.0 " Anterior width of of anal segment, 2.0 " Length of forceps, 7.5 " 1903.] NATURAL SCIENCES OF PHILADELPHIA. 313 A NOTE ON THE COMMON BOTTLENOSED PORPOISE OF THE NORTH ATLANTIC, TURSIOPS TRUNCATUS (Montagu). BY FREDERICK W. TRUE. The common Tursiops of the North Atlantic is generally known m scientific parlance as Tursiops tursio (Fabricius), taking its specific name from the Delphinus tursio of Fabriciiis. During a recent study of Fabricius' species I became convinced that whatever his D. tursio might be it was not the bottlenosed porpoise, so common on the Atlan- tic coast of the United States. The original description by Fabricius has been learnedly discussed by Eschricht, Holboll, Robert Brown, and others with the view of determining if possible what it really represents, but the principal con- cern here is to determine what it does not represent. After describing the head of his D. tursio, Fabricius proceeds as follows: "Teeth in both jaws distant, with obtuse apex, as in Del- phinus albicans [=Delphinapterus leucas], but larger. Body very thick , as in Boops [i.e., humpback whale] and equal with the young of the latter."! Now the teeth of the common bottlenosed porpoise are not far apart, and are acute except in old individuals, and are smaller than those of Delphinapterus. But most significant of all is the remark that the body equals that of the voung of the humpback whale. The young of this whale has a length of from 14 to 18 feet at UrtK which proportions are never reached by the adult bottlenosed porpoise. Taking these facts into consideration it would seem out of the ques- tion to apply Fabricius' name to our common porpoise. The Green- landic name reported by Fabricius for his species was Ncsarnak. Eschricht states that this word means simply "resembhng the Nisa." msa is a name for porpoise adopted by the Greenlanders from the Scandinavian colonists.^ Capt. Holboll had previously stated that the natives appeared to designate the blackfish, Glohicephala, by this name, but certainly not the porpoise currently known as D. tursio.^ In 1868, Robert Brown, in treating of the cetaceans of Greenland, remarked under the head of Glo hicephala: "There seems httle doubt 1 Fabricius, Fauna Groenlandica, 1780, p. 49. 2 Ann. Sci. Nat., 5e ser., ZooL, I, 1864 p. 209 3 Holboll in Eschricht's Untermch. iiher nord. nallthiere, 1849, p. 190. 314 PROCEEDINGS OF THE ACADEMY OF [March, that this is the Delphinus tursio of Fabricius, as the Eskimo name Nesarnak is applied to the present animal." Finally, to clinch the matter, he states that Montagu's Delphinus truncatus {i.e., the bottle- nosed porpoise) "has never been fomid in Davis Strait."'* I am unable to find reference to any Greenland specimens of Twsiops in the museums of Europe and there are none in the U. S. National Museum. The correct name for the bottlenosed porpoise is probably Tursiops truncatus (Montagu), from the Delphinus truncatus described by Montagu in 1821.^ The specific name truncatus has been employed from time to time by various systematists, including J. E. Gray, but Beddard and other recent compilers have generally made use of the inapplicable tursio. There is a possibility that the Delphinus siculus of Rafinesque, described in 1810, is the same as Tursiops truncatus, in wliich case siculus would have priority, but Rafinesque's description is so insuffi- cient that there will probably always be a difference of opinion as to the identity of his species. Under the circumstances it seems unwise to give it serious consideration. Rafinesque's description (translated) is as follows: "Delphinus siculus. "Body oblong, attenuated posteriorly, bluish above, white below; rostrum short, obtuse; teeth equally obtuse. Remarks: This dolphin is called ' Fera ' in Sicily, and has much affinity with Delphinus feres of Bonnaterre, but that is black and has the teeth unequal, alternately longer and shorter. Both differ from D. phocena and D. delphis, which have the teeth acute, and the last the rostrum also."^ Later writers on the fauna of Sicily, so far as I have observed, do not mention the "fera" among the species of porpoises, but Cams states that the common porpoise, or dolphin, Delphinus delphis, is known to the Sicilians as "fera comune," so that there is a probability that Tursiops truncatus might be known as "fera." The Delphinus feres of Bonnaterre is certainly not the same as Tursiops truncatus, as the skeleton of one was 14 feet long, the skull 22^ inches long and 17 inches broad, the total number of teeth but 40, and the skin entirely black. It was probably a species of Glohicephala. * Arctic Manual, 1875, pp. 91-92, from P. Z. S., 1868, with additions and cor- rections. ^ Montagu, Geo., "Description of a Species of Delphinus which appears to be New," Mem. Wernerian Soc. Nat. Hist., 3, 1821, pp. 75-82, 1 PL ® Rapinesque-Schmaltz, C. S., Caratteri eri di alcuni nuovi gen. et sp. anim. et piant. delta Sicilia, 1810, p. 5. 1903.1 NATURAL SCIENCES OF PHILADELPHIA. 315 ADDITIONS TO THE JAPANESE LAND SNAIL FAUNA-No. VIII. BY HENRY A. PILSBRY. Several new forms of Clausilia have been detected by Mr. Hirase since the pubhcation of No. VII of this series.^ They are chiefly from Shikoku Island, and all belong to groups already well represented in the same regions. A single species herein described from Sado is the first Clausilia to be made known from that island. Section HEMIPHiEDUSA Bttg. {Group of C. awajiensis.) Clausilia sadoensis n. sp. Pi. XIV, figs. 6, 7, 8, 9, 10. Shell slender, attenuate above, brown. Surface somewhat glossy, finely striate, paler and smoother above, hardly more coarsely striate on the last whorl. Whorls 10^ to 11, convex, the last compressed laterally, and having a rounded ridge or varix behind the basal and outer lips. Aperture small; peristome white, reflexed, continuous, the parietal margin erect and free. Superior lamella compressed, marginal, continuous with the spiral lamella or interrupted. Spiral lamella short, not quite reaching the middle of the ventral side. In- ferior lamella deeply receding, straightened within, as long as the spiral lamella. Subcolumella lamellar emerging but w^eak. Principal phca short, latero-dorsal. Upper palatal plica short, in the middle joined to the long, nearly straight lunella, which curves inward slightly toward the lower end. Length 15.5 to 16, diam. 3.5 mm. Length 14, diam. 3.5 mm. Clausilium long, parallel-sided, rounded apically, deeply excised on the columellar side of the filament (figs. 8, 9). Misakimura, Sado. Types No. 84,394, A. N. S. P., from No. 993 of Mr. Hirase's collection. Closely related to C. ischna, especially to the variation of that species which I called var. neptis, but C. sadoensis differs in the stronger ridge behind the hp (barely indicated or not perceptible in C. ischna), and in having the spiral and inferior lamellae of equal length inside, while in 1 These Proceedings for 1902, pp. 517-533. 316 PROCEEDINGS OF THE ACADEMY OF [March, ischna and neptis the spiral lamella penetrates decidedly farther than the inferior lamella. Clausilia shikokuensis var. inokuchiensis nov. Smaller than shikokuensis, orange-brown; whorls 9-^, the early ones attenuate. Length 13, diam. 3.2 mm. Inokuchimura, Tosa; No. 84,391, A. N. S. P., from Xo. 506c of Mr. Hirase's collection. A small reddish form, very solid and strong, but structurally like the larger brown C. shikokuensis. Clausilia sus n. sp. PI. XIV, figs. l. 2, 3, 4. 5. Shell quite obesely fusiform, attenuate above, orange-brown, solid and strong, somewhat glossy, finely striate. Whorls 8^, moderately convex, the last compressed laterally. Aperture only shghtly obhque, the peristome expanded and reflexed, somewhat thickened, orange- brown or whitish, a little emarginate at the .position of the superior lamella. Superior lamella compressed, subvertical and marginal, continuous with the spiral lameha, w^hich penetrates to a point above the superior lamella. Inferior lamella deeply receding, thick below, penetrating as deeply as the spiral lamella. Subcolumellar lamella wholly immersed or very weakly emerging. Principal plica visible deep in the throat, entering to a lateral position. Upper palatal plica very short and small, almost obsolete. Lunella strong and a little curved inward below, very weak and low above, where it connects with the middle of the upper palatal plica. Length 11.5, diam. 3.7 mm. Length 11, diam. 3.5 mm. Clausihum (PI. XIV, figs. 2, 3) parallel-sided, thin and acute at the apex, excised on the columellar side of the filament. Muya, Awa, Shikoku Island. Types Xo. 84,393, A. X. S. P., from Xo. 1,007 of Mr. Hirase's collection. The obese shape of the shell and shghtly curved lunella, very strong below and weak above, are characteristic, and the acuminate, acute apex of the clausihum equally so. Some specimens are less swollen, length 12, diam. 3.3 mm. It is not closely related to any other species I have seen, but groups wdth C. shikokuensis. {Group of C. validiuscula.) Clausilia aratorum n. sp. PI. XIV, figs. 15, 16, 17, 18. Shell small, slender and thin, light brown and slightly transparent. Surface glossy, coarsely sculptured with strong, regular, straight striae. Apical whorls smooth, rather large. Whorls 9, somewhat convex, the 1903.] NATURAL SCIENCES OF PHILADELPHIA. 317 last somewhat compressed, very shortly projecting in front. Aper- ture shghtly obHciue, the peristome continuous, narrowly reflexed, but little thickened. Superior lamella marginal, thin and compressed, separated from the spiral lamella, which extends inwanl to the middle of the front side. Inferior lamella deeply receding, thin and straightly ascending wuthin, penetrating deeper than the spiral lamella. vSub- columellar lamella deeply immersed. Principal plica hardly a third of a whorl long, latero-dorsal. Upper palatal plica short and low, a low, short, nearly straight lunella connected with it. Length 9.5, diam. 2 mm. Length 8.5, diam. 1.8 mm. Clausilium long and narrow, rounded distall}-, excavated on the columellar side of the filament. It is very similar to that of C. cenea, but tapers less distally (figs. 17, 18). Shimohanyama, Tosa, Shikoku Island. Types No. 84,438, A. N. 8. P., from No. 1,013 of Mr. Hirase's collection. This small, thin, rib-striate species seems by its palatal armature to be related somewhat to C. micro-peas, but in all other respects it re- sembles the slightly larger C. cenea which was found with it. Both may be more nearly related to C. tosana than to other described species. The lunella in C. aratorum is very weak except close to the upper palatal plica, where it becomes stronger; and in some shells only this stronger portion is present, making with the plica an irregularly triangular callous, without the extension downward shown in the figure. I can see no trace of a lower palatal plica. The subcolumellar lamella has no spiral extension within, on the ventral side of the penult, whorl parallel to the spiral and inferior lamellae, in either this species or C. a'nea. Clausilia senea n. sp. PL XIV, figs. 11, 12, 13, H. Shell slender, fusiform, brown with a bronzed luster, the surface closely, stnjngly striate. Whorls 9J, convex, the last somewhat com- pressed laterally. Aperture slightly oblique, piriform, the peristome thin, free, expanded and slightly reflexed, somewhat emarginate above. Superior lamella thin, compressed, vertical, almost separated from the spiral lamella, which penetrates to the middle of the ventral side. Inferior lamella deeply receding, thin and straightly ascending within, as long as the spiral lamella. Subcolumellar lamella deeply immersed. Principal plica short, dorso-lateral. Upper and lower palatal plicae short, parallel, connected by a low ridge upon which there are three low tubercles. Length 12.5, diam. 2.5 mm. Length 11, diam. 2.3 mm. 318 PROCEEDINGS OF THE ACADEMY OF [March, Claiisilium (PI. XIV, figs. 11, 12) long and narrow, widest and pro- jecting a little on each side near the filament, tapering distally ; excised on the columellar side of the filament. Shimohanyama, Tosa, Shikoku. Types No. 84,392, A. N. S. P., from No. 1,014 of Mr. Hirase's collection. Similar to C. aratorum in sculpture, aperture and lamellre, but having a distinct lower palatal plica and a low lunella upon which several low plicae are seen, in this respect resembling C. tosana. The latter is a more finely striate shell with the last whorl projecting more in front, and the clausilium of a different shape. (Group of C. 'platyauchen.) Clausilia platyderula n- sp. Shell similar to C. platydera, but much smaller; brown; finely and sharply striate. Superior lamella compressed, marginal, continuous with the spiral lamella, which penetrates past the ventral side. The inferior lamella penetrates equally far. The subcolumellar lamella emerges at least slightly. The lunella is like that of platydera. Length 17, diam. 3.8 mm. Length 15, diam. 3.3 mm. Aki, Tosa, Shikoku Island. Types No. 84,443, A. N. S. P.,, from No. 434d of Mr. Hirase's collection. It will be figured in a general review of the species of the platyauchen group, in preparation. Section TYRANNOPH.EDUSA Pils. Clausilia bilabrata var. tosaensis nov. A small form with the shell very strong, the spire entire, and the aperture comparatively small, its greatest length contained 4^ times in that of the shell ; superior and spiral lamellae not united or but weakly so. Whorls about 12. Length 20, diam. 4.3 to 4.8 mm. Shiujomura, Tosa, Types No. 84,378, A. N. S. P., from No. 1,012 of Mr. Hirase's collection. Section STEREOPHiEDUSA Bttg. Clausilia ecMgoensis n. sp. Shell wide below, rapidly tapering upward, attenuate near the apex, several early whorls hardly increasing in diameter, making the outlines concave above; dark chestnut colored, yellow or paler Ijelow the suture. Surface very glossy, sculptured with coarse, oblique ribs, narrower 1903.] NATURAL SCIENCES OF PHILADELPHIA. 319 than their intervals, becoming finer and then disappearing on the early whorls; the intervals between the ribs more or less crinkled. Whorls 11, moderately convex, the last half of the last whorl flattened and compressed, rounded beneath. Suture marked with a fine white line. Aperture nearly vertical, dark inside, the peristome white, narrowly reflexed. Superior lamella small and low, narrow, not con- tinuous with the spiral lamella, barely reaching the margin or not reaching it. Spiral lamella short, not quite reaching the middle of the ventral side. Inferior lamella forming a strong fold in the aperture, becoming subhorizontal, strongly spiral within, a little longer than the spiral lamella. Subcolumellar lamella barely emerging or immersed. Principal plica quite short, latero-dorsal. Upper palatal plica lateral, short. Lower palatal plica shorter, oblique. No lunella. Length 28.7, diam. 7.7-8 mm. Clausilium wide, thickened and somewhat angular at the apex, somewhat excised on the columellar side of the filament. Myokozan, Echigo. Types No. 84,444, A. N. S. P., from No. 986 of Mr. Hirase's collection. Belonging to the C. japonica group, this species seems well distin- guished by its robust shape and very strong sculpture. It will be figured with the other species of the japonica group of Stcrcophadusa^j in a revision of them now in preparation. Clausilia japonica var. perstriata nov. Similar to japonica except in sculpture, the surface being strongly, deeply and closely rib-striate. Kumanogongen, Echigo. Types No. 84,376, A. N. S. P., from No. 986a of Mr. Hirase's collection. Explanation of Plate XIV. Figs. 1, 2, 3, 4. — Clausilia sus n. sp. Typical form. Fig. 5. — Clausilia sus, a more attenuate variation occurring with the preceding. Figs. 6-10. — Clausilia sadoensis, n. sp. Figs. 11-14. — Clausilia cenea n. sp. Figs. 15-18. — Clausilia aratorum n. sp. 320 PROCEEDINGS OF THE ACADEMY OF [March, LIFE COLORS OF PffiCILIA LIMANTOURI, AND DESCRIPTION OF A NEW HEROS FROM MEXICO. BY HENKY W. FOWLER. The Academy has received a small collection of fishes from Tamaiili- pas, Mexico. They were secured January 16 by Mr. S. N. Rhoads, while on his recent visit to Mexico, in the warm waters of the Victoria river near Victoria. This is a small stream tributary of the Rio Soto la Marina, at the foothills of the Sierra Madre Mountains in western Tamaulipas. The material is in excellent condition, so that it is possible to give the hfe colors of the Poecilia, which were unknown. PCECILIID^. 1. Poecilia limantouri Jordan and Snyder. 1900. Poecilia limantouri Jordan and Snyder, Bull. U. S. Fish Comm., p. 129, fig. 10; Rio Tamesoe, near Tampico, Tamaulipas, Mexico (Coll. J. O. Sny- der, No. 6,165, L. Stanford, Jr. Univ. Mus.). Color in alcohol of adult male. Deep olivaceous above, and becoming paler where it extends down along the middle of the side. Posteriorly the dark upper color extends lower. Each scale on the upper surface margined with darker, and the side of the body sprinkled with diffuse dusky specks. Chin and jaws dusky. Lower surface of the head cadmium-yellow, burnished with brassy-orange, and then becoming whitish below. Cheek and lower opercle marked with diffuse dusky specks. Abdomen and lower half of the l^ody pale orange, becoming whitish below, and grayish posteriorly. Five or six longitudinal rows of bright orange spots running along the side, the uppermost following the lateral line, and the others parallel. Dorsal with its basal half dark, the rays deep gray, and the membranes black. Outer j^ortion of the dorsal broadly bright orange. Caudal dark for two-thirds its basal portion, the median basal scales more or less silvery -gray, and then a broad black transverse median band formed by the black membranes, the rays grayish. Outer portion of the caudal broadly deep orange. Pectoral very dilute brown, the base obscurely orange, and then a ])ale slaty blotch. Ventrals and anal dilute orange. Iris deep brown. Considerable ^•ariation occurs, many examples having as many as a PROC. ACAD. NAT. SCI. PHILA. 1903. 1^ c A 10 CRAWLEY ON GREGARINES. PROC. ACAD. NAT. SCI. PHILA. 1903. PLATE 11. CRA^vVLEY ON GREGARINES. PROC. NAT. SCI. PHILA. 1903. PLATE III. CRAWLEY ON GREGARINES. PROC. ACAD. NAT. SCI. PHILA. 1903. PLATE IV. MONTGOMERY ON HABITS OF SPIDERS. PROC. ACAD. NAT. SCI. PHILA. 1903. PLATE V. MONTGOMERY ON HABITS OF SPIDERS. W g CO r G c r > > m t— I d o a w > o a w O w ifil* PROC. ACAD. NAT. SCI. PHILA. 1903. PLATE IX. 11 '- i3 PILSBRY. MOLLUSCA OF WESTERN ARKANSAS, ETC. PROC. ACAD. NAT. SCI. PHILA. 1903. C^-j. Ih ^ 4 a^ ^W" 6a r- n 1 . 8* PILSBRY. MOLLUSCA OF WESTERN ARKANSAS, ETC. PROC. ACAD. NAT. SCI. PHILA. 1903. PLATE XI. 11 11 CL 111 PILSBRY. MOLLUSCA OF WESTERN ARKANSAS, ETC. PROC. ACAD. NAT. SCI. PHILA. 1903. PLATE XIT. WHEELER ON LEPTOTHORAX. PROC. ACAD. NAT. SCI. PHILA. 1903. PLATE XIII. NYCTALEMON CURVATA SKINNER. PROC. ACAD. NAT. SCI. PHILA. 1903 PLATE XIV. ^ i ( \ >,- A ^ /f^'-*^' rc\ \1 13. \m \ 11. 14. 12., \J ^T'^ii 15. 16. 17 18 PILSBRY. JAPANESE LAND SNAIL FAUNA. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 321 dozen distinct vertical dusky-brown bars on the side of tlie back. Other examples show a blackish axillary blotch or several rather large dusky blotches either at the base of the caudal, on the back, or on the abdomen. Some examples are almost black above, and with intense orange on the fins. Nineteen examples. Adult female. Olivaceous above, and extending over the greater part of the side, each scale margined with darker. About six distinct longitudinal series of dusky dots along the side, parallel \\dth the lateral line. Lower surface of the body pale brown or brownish-white, somewhat silvery on the chest. Just above the anal many pale diffuse dusky specks. Dorsal, caudal and pectoral dilute brown, a little deeper basally. Anal with a pale orange tinge. Ventrals whitish. Many have the anal rather bright orange, and dark olivaceous blotches distributed with the same variations as seen with the dark blotches of the males. One female has the three series of dots running parallel to and below the lateral line, orange-ochraceous. Thirty-six examples mostly gravid with ova. A large number of both young males and females show the coloration of the adult female, except the distinct series of dots along the side. The lower surface is almost entirely pale and uniform. Anal pale like the ventrals. The variation of dusky l^lotches is found in most of the examples, even the smallest. This species has been taken also in the Rio Ixtla, at Puente de Ixtla,* Morelos, Balsas and La Antigua.^ CIOHLID^. 2. Heros teporatus sp. nov. Head 2f; depth 2^; D.-XV, 10; A. V, 8; P. i, 14; V. I, 5; scales 29 in a lateral series to the base of the caudal, about 17 in the upper part of the lateral line ; 5 scales between the front of the dorsal and the lateral line, and 12 between the latter and the origin of the spinous anal; width of head 2 in its length; depth of head 1^; snout 3j; eye 3}; interorbital space 3^; fourth dorsal spine 2f ; least depth of caudal peduncle 2J; anal Ij; fifth anal spine 2^-^^. Body elongate, compressed, the greatest depth about the origin of the spinous dorsal, and the back not conspicuously elevated. Upper profile evenl}^ convex, though a little more curved than the lower. Caudal peduncle deep, compressed, and its length about two-thirds its depth. Head rather triangular, compressed, and its depth at the posterior 1 Jordan axd Snyder, Bull. U. S. Finh Cornm., 1900, p. 130. 2 Meek, Field Col. Mu.'^. Pub., 6.5, Zool., Ser. Ill, No. 6, 1902, p. 106. 21 322 PROCEEDINGS OF THE ACADEMY OF [Mai'ch margin of the gill-opening about equal to its length. Both profiles are shghtly convex. Snout conic and rather short. Eye small, well anterior, and high. Mouth small, horizontal, and the jaws about equal. Maxillary small, and not reaching opposite the front rim of the orbit. Lips rather fleshy, and the lower forming a free fold across the mandible. Teeth uniserial, conic, and with a patch of villiform ones directly behind. Nostrils well separated, high, and the anterior midway between the tip of the snout and the front of the orbit. Inter- orbital space convex. Preorbital a little over half the width of the eye. Gill-opening carried forward below the anterior rim of the orbit. ^ Scales rather large, finely ctenoid, and of about even size. Head scaly, except the snout and interorbital space, and about five rows on the cheek. Small scales along the bases of the fins, those along the spinous dorsal forming a low sheath. Lateral line high at first, inter- rupted below the posterior portion of the dorsal, then beginning on the middle of the side and running over eleven scales in a straight line to the base of the caudal. Fourth to seventh spines of dorsal even and highest. Longest dor- sal rays a little longer than the longest spines. Last anal spines longest, and a little shorter than the longest anal rays. Caudal trun- cate, with rounded corners. Pectoral rather long, reaching opposite the origin of the spinous anal. Ventral inserted a little behind the origin of the pectoral and reaching almost to the origin of the spinous anal. Color in alcohol more or less olivaceous-brown, with about nine broad darker vertical bands fading out below. The one at the end of the pectoral with a black blotch, and another black blotch at the base of the median caudal rays. Fins dull olivaceous-dusky, the ventrals somewhat paler. Lower surface of the body pale. Length If inches. Type No. 24,242, A. N. S. P. Victoria, on the Victoria river, a tribu- tary of the Rio Soto la Marina, Tamauhpas, Mexico. Coll. S. N. Rhoads. One example. This species is very close to Heros pavonaceus Gar- man,^ differing in the fin radii, the fewer vertical bands and only having two black blotches on the sides. The first of these is placed at the end of the pectoral just below the lateral fine, and the other at the middle of the base of the caudal. In Heros pavonaceus they are more ^ Bull. Mus. Comp. Zool., VIII, p. 93; from a spring near Monclova (Coll. Dr. Palmer). 1903.] NATURAL SCIENCES OF PHILADELPHIA. 323 or less ocellated and vertically expanded, while in H. teporatus they are rounded black spots. (Teporatus, made somewhat warm; on account of the w^arm springs flowing into the Victoria river where this species was taken.) 324 PROCEEDINGS OF THE ACADEMY OF [April April 7. Mr, Arthur Erwin Brown, Vice-President, in the Chair. Eight persons present. Papers under the following titles were presented for publication : ''Certain Mounds of the Central Florida West Coast/' by Clarence B. Moore. "Certain Mounds of the Apalachicola River," by Clarence B. Moore. April 14. The President, Samuel G. Dixon, M.D., in the Chair. One hundred and thirty persons present. A paper entitled "A new Gurnard from Florida, with Notes on the Colors of some other Florida Fishes," by Henry W. Fowler, was pre- sented for publication. The death of J. Victor Carus, a correspondent, was announced. Mr. Howard W. Dubois made a communication, illustrated by lantern slides, on a reconnaissance of the Rockies of British Columbia while prospecting for the platinum metals. (No abstract.) April 21. Mr. Arthur Erwin Brown, Vice-President, in the Chair. Eight persons present. A paper entitled "Nosema geophile, sp. n., a Myxosporidian Parasite of Geophilus," by Howard Crawle}^ was presented for publication. 1903.] natural sciences of philadelphia. 325 April 28. Mr. Arthur ERT\aN Brown, Vice-President, in the Chair. Eighteen persons present. A paper entitled "A List of Arachnida from Hayti, mth Descriptions of New Species," by Nathan Banks, was presented for pubhcation. The death of Theodore D. Rand, a member, April 24, was annoimced. John V. Shoemaker, M.D., was elected a member. The follo^ving were ordered to be printed: 326 PROCEEDINGS OF THE ACADEMY OF [April, THE MUTATION OF HIBISCUS MOSCHEUTOS L. BY JOHN W. HARSHBERGER. The views of De Vries on the origin of species by mutation rest upon the result of seventeen years of experimentation. This botanist holds that new elementary species arise suddenly without transition or inter- mediate forms, and that these species are constant from the moment of their origin and show no resemblance in their characters to the indi- vidual variations exhibited by the parent type. De Vries believes that mutability occurs only at certain periods, and a species might continue in existence indefinitely without giving rise to new forms. This succinctly states the position of the illustrious Dutch botanist upon the origin of species. The observations here recorded are given with the hope that they may add somewhat to the discussion of the problems opened up by the work of De Vries. Hibiscus moscheutos L., the swamp rose mallow, is found abundantly at Sea Side Park, N. J., where it covers acres of the salt marshes. Here, undisturbed by man, the plants offer special advantages for study. It was noticed that no two plants growing in the open meadow were ex- actly alike. Consequently, a study was instituted to determine, if pos- sible, the character of the divergences. All of the plants studied grew practically under the same conditions, although in certain places the soil of the meadow was wetter than in others. So as to make the observa- tions continuous for a number of years and upon the same plants, stakes were driven into the marsh at the base of each plant collected and studied. These stakes were marked by cutting Roman numerals into a planed-off portion at the top. The plants numbered I-XII grew in a wet portion of the meadow near an open slue. Those designated as XIII to XXV grew in the drier portions of the meadow, excepting XXV, which grew beside a hole filled throughout the season with water. Specimens of these plants were submitted to Prof. De Vries, who pronounced upon them as follows, extracted from a letter dated November 26, 1902: "I have studied the Hibiscus moscheutos with the greatest interest, and beheve, with you, that you have hit at cases of mutation parallel to that of 1903.] NATURAL SCIENCES OF PHILADELPHIA. • 327 CEnothera Lamarckiana." The writer at least hopes that more detailed observations will confirm the encouraging words of Prof. De Vries, However, to know whether the mutabiUty is still working, or whether the period has already come to a close, it is of course necessary to make sowing experiments. This was done last fall at Sea Side Park, where seeds obtained from capsules that had matured were marked by stakes and sown in a portion of the meadow destitute of the rose mallow to determine among the different forms which is the original and mutating one, because it is probable that the others would no more mutate or do so only in a restricted manner. Seeds of the several marked plants were also kept for future sowing. All of the plants studied, with the exception of plants III, IV, V, XII, XVII, were more than of one year's duration. As is well known, Hibiscus moscheutos L. is perennial and persists for a number of years. The doubt might arise in the minds of some botanists that the smaller plants are small because they are juveniles and have not reached full maturity. The size of plants III, IV, V, XII, XVII and the shape of their leaves may be accounted for in this way, but the color of the flowers, leaves and markings can hardly be explained by a reference to the juvenile state, because these plants diverge as widely in appear- ance as the other twenty adult plants do from each other. The adult plants differ from each other, as do the plants of one year's growth, and the inference is, therefore, that the difference in size, mode of branching, size'of leaves, shapes and colors of leaves, character of inflorescences, size and color of the flowers is not dependent upon whether the plant is juvenile or adult, but is due rather to the mutations that they have undergone. Dried plants do not show the peculiarities of structure in as striking a manner as do living plants. In drying, the plants have lost form, and the color of the flowers and leaves has faded out. As the botanist, however, stands in the salt meadow where Hibiscus moscheutos L. abounds and runs his eye over the thousands of plants that are found there, he cannot fail but notice the various striking forms character- ized by habit of growth, size and color of the stem, leaves and flowers, that have been produced, as the writer believes, by the process of mu- tation. One plant is tall and has pure white flowers with bronzed leaves. Another is bright green with rose-pink flowers, while stiU another plant is corymbosely branched and has deep rose-red flowers. These peculiarities are mentioned as they occur in twenty-five plants gathered in the summer of 1902 for comparative study. 328 PROCEEDINGS OF THE ACADEMY OF [April, DESCRIPTION OF A NEW GURNARD FROM FLORIDA, WITH NOTES ON TH£ COLORS OF SOME OTHER FLORIDA FISHES. BY HENRY W. FOWLER. A small collection of fishes has recently been received at the Academy from Dade county, Fla., in the Biscayne Bay region. It was made during February of the past winter. The Academy is indebted to Mr. James Spear, Jr., for this addition to its collections. He was also fortunate in being able to forward the specimens in a nearly fresh state of preservation, so that I can give most of the life colors. OARANGID.^. 1. Selene vomer (Linneeus). 1758. [Zeus] Vomer Linnseus, Syst. Nat., Ed. X, I, p. 266; in America. Color in alcohol with brilliant silvery prevailing, and everywhere with beautiful purplish and bluish reflections. Back and upper surface of the caudal peduncle purplish-brown, deepest along the dorsal profile. A large scattered tract of dusky dots, becoming darker and most tUs- tinct immediately above the eye. Axil of the pectoral pale yellowish with dusky dots. Spinous dorsal dusky. Soft dorsal whitish, with the long anterior rays dusky, the longest ray very dark. All the other fins whitish with more or less dusky on their outer portions. Upper and lower anterior margins of caudal lobes dusky. One example 10 inches in length, from New river. NOMEID^. 2. Nomeus gronovii (Gmelin). 1788. [Gobius] Gronovii Gnielin, Syst. Nat., I, p. 1,205; in oceano ameri- cano zonae torridae. Color in alcohol dark brown on the back and top of the head, the sides and lower surface silvery- white. The dark brown of the back extends down on the sides in the form of five large saddle-like blotches becoming grayish below. Lower side of the body with six or seven grayish-brown blotches. Lower surface of the head with several grayish blotches. Fins slightly tinged with straw-color, which may be due to the preservative. Spinous dorsal black. Soft dorsal ^\ith three broad transverse blackish vertical bands, leaving only narrow pale spaces between. Anal with three narrow transverse blackish 1903.] NATURAL SCIENCES OF PHILADELPHIA. 329 vertical bands, which extend a little on the lower surface of the body, and leave broad spaces between. Two small blackish spots on the lower edge of the caudal peduncle. Caudal with a blackish blotch at the base of each lobe. Median portion of each caudal lobe blackish. Pectoral with a longitudinal median blackish streak, becoming diffuse on the outer portion. Membranes of the ventral black, except the outer portion, which is pale straw-color, like most of the rays. Length 4f inches. One example taken from below a Portuguese man-of-war {Physalia) in Biscayne Bay. SERRANID^. 3. Diplectrum formosum (Linnoius . 1766. fPerca] formosa Linnaeus, Syst. Nat., Ed. XII, p. 488; Carolina (Coll. D. Garden). Color in alcohol pale brown above, the lower surface silvery-white. Back and upper side with eight broad transverse deep browm bands, with alternating deep brown narrow bands between. A deep brown band from the tip of the snout to the base of the upper caudal lobe. Above this two similar narrow dusky-brown bars. Where all these bars cross dark areas are formed. The side of the head and trunk are also marked with narrow pale slaty longitudinal lines, each one with a little darker margin. Dorsals and caudal with dilute yellowish- brown, the former marked on the spinous portion with two broad pale slaty longitudinal bands which run back on the soft fin higher and lower, the intervening space marked with another similar parallel band. All of these bands are darker along their edges, and run liigher on the posterior soft dorsal, where a fourth runs backward from the base of the sixth ray. Caudal with six more or less connected lines or bands, especially above, so that the fin has a spotted appearance. Anal pale yellowish-white, with many dilute slaty linos. Pectoral and ventral whitish. A dusky spot on the chin. Interorbital space with two narrow connecting lines, also one between the nostrils. Length 8 inches. One example from lower Biscayne Bay. LUTIANID^. 4. Neomaenis apodus (Forster). 1792. Perca apoda Forster, in Walbaum, Pet. Arted. Gen. Pise, III, p. 351 (Based on Perca marina pitinis, etc., Catesby, Nat. Hist. Car. Flor. Bah. 11, 1731, p. 4, PI. 4, lower figure; no locality.)^ Color in alcohol more or less faded dull uniform olive, pale below. ^ Credited by Pennant (Arctic Zoology, II, 1792, p. 385) to the Bahama isles. 330 PROCEEDINGS OF THE ACADEMY OF [April, Pale vertical streaks or lines obscure. A slaty line below the eye to the end of the opercle, and another below this. A brown streak across the base of the pectoral. Two j^oimg examples from Card Sound. 5. Ocyurus chrysurus (Bloch). 1797. Sparus chrysourus Bloch, Ichthyologie, III, Pt. 8, p. 25; les eaux du Bresil. Sparus Chrysurus Bloch, I. c, Plate 262. Color in alcohol with upper surface deep ohvaceous-gray, and the sides and lower surface rosy. A bright greenish-yellow band from the tip of the snout, at first narrow, and then widening posteriorly until it includes the entire caudal fin. Above this, back, and on the upper surface of the head, are a number of dark greenish-yellow spots. Each scale on the cheek, and lower surface of the head, with a deep rosy spot. Above the lateral line oblique dusky lines with a rosy tinge running up to the base of the dorsal. Between the arch of the lateral line, and the golden lateral band, are about four narrow, deep rosy lines running parallel. Just below the yellow band are also two parallel deep rosy lines leaving a paler rosy space between, then the rest of the lower surface is marked with seven or eight longitudinal bright greenish-yellow narrow bands alternating with similar ones of deep rosy. The lower edge of the abdomen, and under surface of the head, is whitish. The scales on the chest and lower part of the abdomen are more or less whitish basally, producing an imbricated appearance. The top of the head is marked with bright olivaceous spots. Dorsals and anal ohva- ceous-yellow, the soft dorsal becoming very bright posteriorly. Pectorals and ventrals yellowish- white. Iris red. Lips dusky. Inside of the mouth and gill-opening whitish. One example 7 inches long, from lower Biscayne Bay. Comparison with Brazilian examples is desirable. 6. Haemulon sciurus (Shaw). 1803. Sparus Sciurus Shaw, Gen. ZooL, IV, p. 439, PI. 64; American seas. Color in alcohol ohvaceous-dusky above, paler on the sides, and be- coming yellow below. Everywhere more or less brassy. Side wdth about ten longitudinal slaty-blue stripes, each one edged narrowly with deep slaty. Spinous dorsal dilute grayish-green. Soft dorsal dusky. Caudal dusky, the outer portion dilute greenish-yellow. Anal bright yellowish-green with the base dusky. Pectoral grayish. Ventral bright yellowish-green. Each ramus inside the mouth JDright brick-red. Inside of the gill-opening reddish above. Peritoneum black. One example 9^ inches long, from lower Biscayne Bay. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 331 7. Haemulon plumierii (LactJpMe). 1800. Labrus Plumiqrii Lacepede, Hist. Nat. Poiss., Ill, pp. 432, 480, Plate 2, fig. 2; Amerique. Color in alcohol brassy grayish-brown, darker on the upper surface. Side with many Ught bkie Unes, each one bordered with deep slaty, mostly longitudinal, and those running from the pectoral and ventral very oblique. Fins dusky, and the margins of the membranes of the spinous dorsal narrowly deep brown. Base of ventral dilute greenish. Inside of the base of pectoral yellowish with bluish lines. A brown bar across base of the pectoral. Inside of the mouth, and lower lip posteriorly, orange-red. Outer portion of first ventral ray whitish. One example 7^ inches long, from lower Biscayne Bay. SPARID^. 8. Calamus bajonado (Schneider). 1801. [5parus] Bajonado Schneider, Syst. Ichth., p. 284. (Based on Bajo- nado Parra, Descr. Piez. Hist. Nat., 1787, p. 13, PI. 8, fig. 1; Cuba.) Color in alcohol brassy-brown, the margin of each scale with a nar- row darker submarginal border. Back, a little deeper than the lower surface, which is washed with silvery. A dull bluish line below the eye running forward toward the tip of the snout. Below this an in- distinct bluish line, and also several bluish spots on the cheek. A bluish line across the upper edge of the opercle. Two narrow bluish lines running out from the front of the eye. A dusky band from the lower rim of the orbit to the corner of the mouth, and across the mandi- ble. Vertical fins, and ventral, dull slaty mottled with brownish. Pec- toral pale brownish-white. Axil of pectoral pale yellowish. Perito- neum silvery. Iris flesh-colored, upper anterior portion dusky. One example 7^ inches long, from lower Biscayne Bay. 9- Lagodon rhomboides (Linuteus). 1766. [Spams] rhomboides Linnteus, Syst. Nat., Ed. XII, I, p. 470; in America (Coll. D. Garden). Color in alcohol dusky-olivaceous on the back, merging into silvery- white on the lower surface. About seven vertical dusky cross-bands, with as many narrower and indistinct ones between. A deep brown blotch on the lateral line above the base of the pectoral. Side with eleven longitudinal slaty-blue bands, the intervening spaces more or less gilded. Fins, with the exception of the ventrals, pale dusky, the basal portions of the dorsals with many large gilded spots. Anal more or less pale golden. Ventral whitish with a large golden basal blotch. Iris clouded with dusky. One example 4f inches long, from New river, Fla. 332 PROCEEDINGS OF THE ACADEMY OF [April, GERRID^. 10. Eucinostomus harengulus Goode and Bean. 1879. Eucinostomus harengulus Goode and Bean, Proc. U. S. Nat. Mus., p. 132; West Florida (Coll. MM. Kaiser and Martin). Color in alcohol very pale olivaceous-brown above, more or less sil- very, and the lower surface silvery- white. The back is also clouded with a deeper shade of the general body color. All of the scales with a more or less bright metallic bluish or purplish luster. Snout brown- ish. Upper portion of spinous dorsal blackish. Upper vertical fins slightly brownish, other fins whitish. Iris silvery, with dusky blotches. One example 3|^ inches long, from New river. 11- Xystsema cinereum (Walbaum). 1792. Mugil c/nerc ;;.? Walbaum,= Pet. Art. Gen. Pi.sc, III, p. 228. (Based on Turdus cinereus peltatus Catesby, Nat. Hist. Car. Flor. Bah., II, 1731, p. 11, PI. 11, lower figure; no locality.)^ Color in alcohol pale olivaceous-brown, the lower surface whitish. Body washed with silvery, sho-^ing purplish-blue reflections. Side with seven vertical dull purplish bands running down from the back. Vertical fins with minute dots of dusky, also on the ventrals. Spinous dorsal and ventrals washed with dilute greenish-yellow. Iris grajdsh- silvery. Peritoneum white. One example 11 f inches in length, from New river. LABRID^. 12. Iridic bivittatus (Bloch). 1797. Labrus bivittatus Bloch, Ichthyologie, III, Pt. 8, p. 107, PI. 284, fig. 1 ; j'eu ignore la patrie. Color in alcohol greenish-olivaceous, pale or whitish below, and the edge of each scale more or less tinted with dull coppery. A dull purplish band running from the upper part of the eye, narrowly bor- dered on each side, while on the head, with blue. It runs along the base of the dorsal. A dusky purplish band runs from the side of the snout through the eye to the base of the caudal above, and along the side of the trunk it is very broad. A similar narrow band from the pectoral along the lower side. A narrow pale dusky band starts above the opercle and includes the lateral line till it descends posteriorly. '^ I have recently rejected Walbaum's name Raia birostris, adopting Raia manatia of Schneider. A re-examination of Walbaum convinces me that the word birostris is evidently a typographical slip, and that sucli names as he used in the " Additamentum" are availaljle as binomials, for in every case a specific name is italicised. The name Mobulidce should supersede Mantidce. See Science, XVII, April 10, 1903, p. 594. ^ Pennant says {Arctic Zoology, II, 1792, p. 377), under his discussion of Labrus /imi«Za Linnaeus, that it inhabits the seas of Carolina, "as I suppose the three former do," of whicli the one immediately preceding is Catesby's fish. 1903.] NATURAL SCIEXCES OF PHILADELPHIA. 333 A pale purplish streak below the base of the pectoral. A purpUsh Y-shaped bar on the opercle, with narrow pale blue margin, and with a deep violet-blue spot above. Upper Up oUve-green. Cheek pale. Mandible with two brownish-purple cross-bars narrowly bordered with pale bluish. Dorsal fins dilute orange, basally dusky, with greenish spots narrowly margined with pale bluish. Anal similar, grapsh, and with dull rosy tints. Caudal deep orange-red, marginal portion of the fin blackish, and with pale oblique bluish stripes. Axil of pectoral pale green. Iris bright orange edged with green. One example 8f inches long, from lower Biscayne Bay. I am not certain that my fish is identical with Bloch's account, as it also does not agree with his figure. Lacepede's account is less satisfactory. TETR AOD ONTID^ . 13. Cheilichthys testudineus (Linnseus).* 1758. [Tetraodon] testudineus Linnaeus, Sj'st. Nat., Ed. X, p. 332; in India.^ Color in alcohol dark slaty-brown, the sides gradually becoming whitish below with a dull ochraceous tinge, especially along the lower side of the head. Large deep blotches of blackish-brown on the back, made up of clusters of crowded spots which leave winding paler lines between. On the sides these spots become separated and are very distinct against the paler background. Axil of the pectoral brown. Fins dull dusky with an indistinct tinge of dilute olivaceous, except the anal which is white. Lower surface of the body chalky-white. Iris yellowish-gray. One example 8f inches in length, lower Biscayne Bay. TRIGLID-ffi. Merulinus Jordan and Evermann. Bull. U. S. Nat. Mus. (Fish. N. Mid. Amer.), No. 47, II, 1898, p. 2, 148. Type [Trigla] carolinn'^ Linnscus, Mantissa Plant., 1771, p. 528; in Carolina' marl (Coll. D. Garden). 14. Merulinus salmonicolor sp. nov. Head 3; depth 4|; D. X-I, 11; A. II, 9; P. Ill, 12; V. I, 5; scales * Cheilichthys is given generic rank for those American forms with a broad, flattened, or "only slightly concave interorbital space. Type Tetraodon testu- dineus Linnseus. ^ The type of the species probably came from tropical America. The rude accounts of Clusius and Willughby throw little light on the subject. I have been unable to consult Dr. Gunther's reference, " Tetrodon testudineus, L. Amoen. Acad., i, p. 309, tab. 14, fig. 3." The edition of Amoenitates Academicse, etc., in the Academv's librarv, has an account of Ostracion oblongus glaber, etc., by Balk (1749, p. 591).' It is based on Artedi, Willughby, Clusius, Ray and Sloan, and no locality is given. If, however, Linnaeus sliould be found untenable, Tetrodon geometricus Schneider is the next name available with certainty. ^ This name is transposed from the original. 334 PROCEEDINGS OF THE ACADEMY OF [April, 95 in the lateral line to the base of the caudal, and about 6 more on the base of the caudal; 10 scales between the origin of the soft dorsal and the lateral line in a vertical series, and about 24 between the latter and the origin of the anal; pores 50 in the lateral hne to the base of the caudal; width of the head If in its length; depth of head Ijjj: snout 2; eye 5f ; maxillary 2^; mandible 2^; width of mouth at cor- ners 2f; interorbital space 5^; first dorsal spine 2f ; third 2^; first dorsal ray 2J; first anal ray 3^; eighth anal ray 2|f; upper caudal lobe 1^; least depth of caudal peduncle 4^; ventral 1^; space between bases of innermost ventral rays 3f ; ventral spine 2f. Body rather stout, cylindrical, short in comparison, and thickest anteriorly, so that the greatest depth falls opposite the fourth dorsal spine. The upper profile is evenly convex from above the eyes to the upper region of the caudal peduncle. Lower profile more or less straight. Chest and breast flattened, and broad. Caudal peduncle and posterior portion of the trunk compressed, so that the depth of the former, measured from the base of the last dorsal ray, is about one and four-fifths in its length. The upper and lower surfaces are also somewhat fiattened. Head moderately large, broad and high, and its lower profile hori- zontal. Snout long, broad, obhque, and the profile straight. The upper jaw projects down a little beyond the anterior profile of the snout, which, when viewed from above, is seen to be concave. On each side broad round processes with denticulated edges are developed, and pro- ject anteriorly. At the posterior outer edge of each is a small spine. Eye a little posterior in the head, and impinging on its upper profile. A furrow runs from the lower edge of the eye down along the front of the snout to the nostrils. ^louth large, the jaws moderate, and the lower inferior. Maxillary falling a httle short of the front rim of the orbit, and its distal expanded extremity about two-thirds the orbit. Teeth minute, and in rather broad, rasp-like bands, in the jaws. Sim- ilar but smaller teeth in narrow bands on the vomer and palatines. Tongue large, broad, thick, smooth, and only the small tip free in front. Lateral articular region of the mandible with a short series of small serrse. Nostrils anterior, near together, the anterior with a little flap, and the posterior a little nearer the tip of the upper jaw than the eye. Interorbital space rather narrow, a trifle less than the inter- nasal space, flattened in the middle, and each side a httle elevated. Top of the head very slightly convex. Bones of the head all with fine radiating strise. Four preocular spines, the innermost the larger. Two postocular spines. Two tym- 1903.] NATURAL SCIENCES OF PHILADELPHIA. 335 panic spines, the posterior rather large. A small coronal spine. Nuchal spine long. Preopercular, opercular and humeral spines large and strong, and of more or less even size. Anterior and lateral labial bones each with a small spine, that of the latter minute. Gill-opening extending forward nearly opposite the posterior nostril. Gill-rakers in 1 + 8 vi, compressed, short, their inner edges with minute asperities, and the longest a Uttle shorter than the longest filaments, which are half the eye diameter. Pseudobranchise large. Isthmus very broad, flattened, and the gill-membrane free. Peritoneum silvery. Anus close to the origin of the anal fin. Scales small, finely ciliated, and greatly reduced on the front of the back. Scales on the belly small and smooth. Base of the caudal scaly. Space between pectoral and ventral naked. Lateral line com- posed of simple tubes, high, continuous, mostly concurrent with the dorsal profile, and running down a httle above the middle of the caudal peduncle. Origin of the spinous dorsal directly after the base of the pectoral, the third spine the highest, the first longer than fifth, and the others all graduated down to the last, which is very short. The soft dorsal is inserted a little nearer the front of the eye than the base of the caudal, the anterior rays a little the liighest, and like the spinous fin, depress- able in a rather broad groove. Margin of the soft dorsal with small notches. Caudal long, truncate. Anal inserted below the origin of the soft dorsal, the spines flexible, shorter than the rays, and the second a little the longer. Margin of the rayed anal deeply notched, and the penultimate ray the longest. Pectoral long, reaching opposite the base of the last anal ray, and the eighth ray the longest . Ventral long, reaching the anal fin, and the third and fourth rays the longest and about equal. Color in alcohol deep olivaceous-brown on the upper surface, indis- tinctly variegated with small darker specks. On the head these are more distinct and form short wavy lines. Edges of the mouth and gill-opening pale yellowish, the latter somewhat tinged with salmon. Lower surface of the body white, the flanks tinted with salmon-pink. Anal pink. Dorsals and caudal dusky, the former marked with diffuse darker spots. A dark spot in front of the third dorsal spine near its margin. Pectoral blackish, variegated with rather large black spots, and its inner ray whitish basally. Ventrals whitish, with many dusky dots. Iris grayish-brown. Inside of the gill-opening brassy-dusky and blackish. Length 10 inches. 336 PROCEEDINGS OF THE ACADEMY OF [April, Type No. 24,343, A. N. S. P. New river, Dade collnt3^ Fla. Coll. Mr. James Spear. One example. This species is related to Merulinus carolinus, M. scituhis, M. roseus and M. alatus. It is closer to the latter in squama- tion, but differs in more gill-rakers and shorter pectoral. The pre- operciilar spine is also shorter. From the other species it is chiefly distinguished by its bright salmon-pink anal fin. (Sal)no, the salmon, color, color; referring to the anal fin.) 15. Prionotus evolans (Linnajus). 1766. [Trigia] evolans Linnaeus, Syst. Nat., Ed. XII, p. 498; in Carolina. Color in alcohol dull olivaceous-dusky above, the lower surface whitish. Back \Adth three dark cross-bands, and indistinct darker mottlings. Head also mottled with darker, sometimes forming diffuse lines. A blackish bar across the interorbital space. Dorsals and caudal grayish-brown diffused with darker. Pectoral slaty-black with rather large round blackish spots and the edge of the inner ray whitish. Other fins white, the ventrals soiled a little on their median outer sur- face. Inside of the gill-opening dull yellow, reddish on the shoulder- girdle. Eye deep gray-brown. One example 9f inches long, from New river. ECHENEIDID^. 16. Echeneis alba-cauda MitchiU. 1817. Echeneis alba-cauda Mitchill, Amer. Month. Mag. Crit. Rev., II, p. 244; in the bay of New York, June 22, 1815. Color in alcohol dull brown. A broad lateral l^lackish-brown band running from the mouth to the base of the caudal, and bordered nar- rowly below wdth dull slaty-white. Lower rays of the dorsal and anal dusky-brown, and this color extending forward on these fins for the same height, but becoming a little pale. The remaining outer portions of the anterior rays and edges, narrowly, of the others, whitish. Caudal blackish, with the upper and lower corners narrowly whitish. Yen- tral and pectoral blackish, with grayish posterior margins. Iris blacldsh-brown. Length 5f inches. One example taken off of a large shark, evidently Scolidon terrce- novcel, judging by a photograph, captured in Carson's creek at Bis- cayne Bay. It has 22 laminae in the disk, and the median caudal rays project. I prefer the separation of this species from the Indian, until examples are compared. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 337 NOSEMA GEOPHILI, sp. n., A MYXOSPORIDIAN PARASITE OF GEOPHILTJS. BY HOWARD CRAWLEY. On May 21, 1900, a specimen of Geophihis was taken in the Harvard Botanical Garden, at Cambridge, Mass. The intestine of this centi- pede was teazed and crushed upon a slide, fixed in corrosive subUmate and stained with thionin. The preparation showed innumerable individuals of the vegetative stage of a coccidian, probably a species of Eimeria, and some 30-40 specimens of the parasite here figured and described. Of these, the smallest, shown in figs. 1 and 2, were for the most part nearly oval cells, with occasionally a blunt prolongation at one end. They ranged in length from 30 microns upward. There was no distinc- tion to be made out between ectoplasm and endoplasm, the cell sub- stance being essentially uniform. It stained rather deeply with thionin, was dense in structure and beset with numerous vacuoles. No definite membrane could be made out. These smaller bodies were mostly uninucleate, although in some the nucleus had already divided. The nuclei themselves were ellipsoidal bodies, with a faintly staining ground substance and a large and conspicuous karyo- some. The ground substance was doubtless a liquid in the living animals. The karyosomes stained intensely in thionin and were in almost every case vacuolated. The larger specimens, shown in figs. 3 and 4, attained a length of 1.50-200 microns. Their protoplasm was much less dense than that of the smaller forms, and showed a spongy structure. There was no differentiation into ectoplasm and endoplasm. The nuclei were gen- erally arranged in pairs, indicative of recent division, but the prepara- 22 338 PROCEEDINGS OF THE ACADEMY OF [April, tion showed none in the actual process. The nuclei of both large and small organisms were essentially alike. The irregidar form and multi-nucleate condition of this sporozoan places it in the My^sosporidia. Its occurrence as a free-living form in eitherjthe body-cavity or intestine of an arthropod seems to warrant placing it, at least provisionally, in the genus Nosema. The specific name geophili is appropriate in view of the host. The observation is interesting in view of the fact that this is the first recorded case of a myxosporidian being found in a myriapod. It is also worthy of note that while Geophilus is thus shown to be parasitized by representatives of three orders of Sporozoa, infection is only occa- sional. Leger, describing Rhopalonia geophili, a gregarine, says that it is very rare and I have never yet encountered it. I have seen the unidentified coccidian mentioned above in one or two other hosts, but the myxosporidian in only this one case. This comparative immunity is doubtless due to the solitary habit of the centipede. 1903.] natural sciences of philadelphia. 339 May 5. The President, Samuel G. Dixon, M.D., in the Chair. Thirty-nine persons present. A revision of the By-Laws was adopted providing that hereafter (Chap. XII, Article 1) "the stated meetings of the Academy shall be held on the first and third Tuesdays of each month from October to May inclusive, at such hour as may be fixed by the Council ; but no change shall be made in the hour except after one month's notice." Also that (Chap. VII, Art. 4) "the Committee [on Pubhcation] shall examine all papers intended for publication and shall print those ap- proved by it as far as practicable in the order in which they were pre- •sented. It shall also prepare and print in the Proceedings abstracts from the minutes of the meetings of the Academy and summaries of verbal communications made to the Academy or any of its Sections." The following were ordered by the Publication Committee to be printed : 340 PROCEEDINGS OF THE ACADEMY OF [May, A LIST OF ARACHNIDA FROM HAYTI, WITH DESCRIPTIONS OF NEW SPECIES. BY NATHAN BANKS. During the spring of 1899 Mr. R. J. Crew collected some spiders in Hayti, mostly from the vicinity of Port-au-Prince. Comparatively few spiders have been recorded from this island, and many of these are from other parts and are large spiders. According to my instructions Mr. Crew collected the smaller species, and thus secured a number of new forms, as well as others unknown from the region. In all 63 spiders and 3 Phalangids were taken. The Epeiridce with 17 species and the Attidse with 15 species are more fully represented than the other families. Most of the spiders are known only from the West Indies, but about 17 species occur also in the southern United States. The follomng is the list: Araneida.- Scytodes longipes Lucas. Scytodes fusca Walck. Nops coccineus Simon. Pholcus tipuloides Koch. Widfila parimla Banks. Chiracanthium indusum Hentz. Aysha tenuis Koch. Anyphcena gracilipes Banks. Hypsijiotus humilus Keyserlins. Trachelas hicolor Keyserling. Eutichurus insulanus Banks. Theridium studiosum Hentz. Theridula sphcerula Hentz. Theridula triangularis Keyserling. Floronia coccineus Hentz. Argyrodes argyrodes Walck. (tiephilce Tacz.). Argyrodes trituherculatus Becker. Ulohorus geniculatus (31ivier. Miagrammopes sp. (immature), Tetragnatha sp. (female). Argyroepeira argyra Walck. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 341 Argyroepeira bigibhosa Keyserling. Alcimosphenus licinus Simon. Epeira theisii Walckenaer. Epeira circulata Walck. {halaustind%iQ,Co6k) . Epeira fusco-vittata Keyserling. Epeira undecim-tuberculata Keyserling. Epeira foliata Hentz. Eustala prompta Hentz. Singa crewii Banks. Cyclosa walckenaeri Keyserling. Wagneria tauricornis Cambridge. Acrosoma armatum Fabriciiis. Gasteracantha hilaris Thorell. Gasteracantha tetracantha Linnaeus. Gasteracantha cancriformis Walck. Gasteracantha sexserrata Walck. Misumena asperatus Hentz. Misumena sp. (immature). Stephanopis rugosa Taczanowski. Isaloides toussaintii Banks. Olios antiguensis Keyserling. Heteropoda venatoria Linnaeus. Selenops insularis Keyserling. Dolomedes -marginellus Koch. Oxyopes salticus Hentz. Peucetia viridans Hentz. Hamalatiwa grisea Keyserling. Wala vernalis Peckham. Wala peckhamce Cockerell. Plexippus paykulli Aud. and Sav. Metacyrba pictipes Banks. Nilakantha cocker elli Peckham. Prostheclina perplexa Peckham. Prostheclina morgani Peckham. Prostheclina signata Banks. Pellenes locuples Simon. Mogrus cephalotes Simon. Cybele haytiensis Banks. Icius separatus Banks. Synageles sp. (immature). Lyssomanes viridis Hentz. Lyssomanes antillianus Peckham. 342 proceedings of the academy of [^i^y, Phalangida. Cynorta obscura Banks. Cynorta v-album Simon. Stygnus insidanus Banks. Descriptions of New Species. Wulfila parvula n. sp. Cephalothorax pale yellowish, black around eyes; mandibles red- brown; sternum yellowish; legs and abdomen paler, unmarked. The posterior eye-row procurved, P.iM.E. fully diameter apart, about as far from the rather smaller P.S.E. ; A.M.E. quite small, once and one- half their diameter apart, scarcely as far as the larger A.S.E., the latter close to and plainly smaller than the P.S.E. , quadrangle of M.E. higher than broad, and broader behind than in front. Mandibles quite large; in the male very much elongate, nearly as long as femur I, tapering and divergent beyond middle, porrect, bearing a row of teeth below. Legs long and slender, three pairs of spines under tibiae I and II, two pairs under these metatarsi. Male palpus long, on the upper side of the tibia is a row of short black spines, and a group of similar spines at the tip. Length, d^ 3.2 mm.; 9 2.8 mm. Several specimens of both sexes. Anyphsena gracilipes a. sp. Pale yellowish throughout; dorsum of abdomen rather more grayish, with a few indistinct elongate gray streaks above; eyes on black spots. Eyes subequal in size, except those of the anterior median pair which are much smaller than the others. Posterior eye-row procurved; P.M.E. fully diameter apart, nearly as far from the equal P.S.E.; A.M.E. about diameter apart, very small, a little farther from the much larger A.S.E. Mandibles moderately long, but not porrect. Legs extremely long and slender, the anterior tibia nearly twice as long as the cephalothorax ; the tibia and metatarsi I and II have each two pairs of very long spines beneath. Abdomen quite slender; the ventral furrow much nearer to the vulva than to the spinnerets. The epigynum shows two dark almond-shaped cavities, their pointed ends toward each other and curved backward. Length 6 mm. One specimen, female. Singa crewii n. sp. Cephalothorax pale yellowish, with a median blackish stripe, darker around eyes. Mandibles yellowish-brown; sternum pale, sides brown- 1903.] NATURAL SCIENCES OF PHILADELPHIA. 343 ish. Legs pale yellowish, tips of tibise dark; palpi pale on base, black- ish beyond. Dorsum of abdomen pale grayish- white, with a broad gray folium, distinct in outline, but pale within, its anterior edges broadly marked with black, this black in front connects to the basal ends of the black side stripes, these stripes are broad black ones on each upper side of abdomen, below them the sides are pale yellowish; on the middle of the venter is a large black area, containing a central square pale spot. It is of the usual structure of the genus, the legs, however, rather longer than usual. Length 4 mm. One female; a very pretty species. Isaloides toussaintii n. sp. Cephalothorax pale yellow, more brownish on the sides; mandibles, palpi and sternum pale yellowish; hind two pairs of legs pale yellow; anterior pairs more red-brown, with a pale spot beyond middle of tibia; abdomen pale gray, with a broad brown stripe from base to tip, with irregular sides, and covering the tubercles ; venter pale, rather darker in the middle. Cephalothorax flat above, eye-region but little ele- vated; clypeus low; no dorsal' groove ; posterior row recurved, P.M.E. about one and one-half diameter apart, and fully as far from the P.S.E. ; A.M.E. much lower down than A.S.E. so that the eyes are really in three rows, A.M.E. not half the size of P.M.E., but fully twice the diameter apart; A.S.E. fully twice as large as P.M.E. and about two diameters apart. Four pairs of spines under tibiae I and II, three pairs under these metatarsi. Length 6 mm. Several specimens ; diflfers from I. octoperlata Sim. in the shape of the vulva, as well as in minor characters. Metacyrba pictipes n. sp. Cephalothorax dark red-brown, a narrow white side-margin; abdo- men greenish-brown, with a pair of parallel submedian pale lines above, made up of dots, rather indistinct; sternum dark red-brown; venter greenish-brown, with indistinct pale lines each side ; tips of palpi white ; mandibles reddish-brown; yellow hairs around eyes and on clypeus. Legs brown, tarsi white, a pale spot beneath on trochanters, patellae above are almost wholly pale yellowish, an indistinct pale line above on femora and tibiae, scarcely visible on the anterior pair. Structure similar to M. tceniola Hentz. Length 3.8 mm. 344 PROCEEDINGS OF THE ACADEMY OF [May, One male specimen, closely related to M. tceniola, but the male palpus has but one long projection on the tibia, the other is quite short; the legs are very differently marked than in M. tceniola. Cybele haytiensis n. sp. Cephalothorax reddish-yellow, with a median white spot with an extension each side and behind; eyes on large black patches; clypeus pale; mandibles pale, with a brown band across middle; sternum and legs pale, the latter with some small brown patches at base of joints and on the tibia near middle; abdomen pale, with gray-brown dots and an irregular stripe each side; venter pale. Male similar, but the side stripes from the median pale mark are indistinct; the anterior pair of legs is infuscate ; the mandibles are brown ; the venter is black ; and there is a black stripe on the ventral side of each coxa. The epigynum shows two large circular marks quite close to each other, and in front two divergent curved lines; the male palpus is quite simple, the tibia is long and with a long, straight projection on the outer side at tip. Length, ? 7 mm., c? 5.5 mm. A few specimens of both sexes. Evidently related to C albopalpis Peck., from Jamaica. Icius separatus n. sp. Cephalothorax yellowish-brown, clothed with white and yellowish scales and scattered long black hairs; eyes on black spots, and a trans- verse black patch in the middle of the eye-region. Mandibles red- brown; sternum pale; legs very pale, the anterior pair more yellow, and its joints tipped with blackish, other legs unmarked. Dorsum of abdomen pale gray, with several large median brown spots forming an incomplete median stripe; each side are many small brown spots; the sides of the abdomen with brown lines ; the venter pale ; the spinnerets reddish. The cephalothorax is rather low and flat, and cpite short; the eye- region is plainly broader behind than in front, and about once and a third broader than long; the A.S.E. are as high as the A.M.E., the latter very large and touching; the eyes of second row are nearer to lateral than to dorsal eyes. Leg I, which is rather longer than leg IV, is much enlarged, five stout spines on femur I, three pairs under tibiae I and II, two pairs under same metatarsi; metatarsus III spined only at tip; anterior coxae separated by width of lip; abdomen short and flat. Length 3.2 mm. One female specimen. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 345 Explanation of Plate XV, Fig. 1. — Cybele haytiensis. Fig. 2. — Wulfila parvula, vulva. Fig. 3. — Isaloides toussaintii, vulva. Fig. 4. — Wulfila parvula, palpus. Fig. 5. — Icius separatus, eyes. Fig. 6. — Anyphcena gracilipes, eyes. Fig. 7. — Metacyrha pictipes, palpus. Fig. 8. — Singa crewii, vulva. Fig. 9. — Cybele haytiensis, vulva. Fig. 10. — Anypha;na gracilipes, vulva. Fig. 1 1 . — Icius separatus, vulva. Fig. 12. — Cybele haytiensis, palpus. 346 PROCEEDINGS OF THE ACADEMY OF [^lay, NOTES ON A FEW FISHES FROM THE MOSQUITO COAST OF NICARAGUA. BY HENRY W. FOWLER. A few fishes have recently been received at the Academy of Natural Sciences of Philadelphia from the Rev. William H. Fliick. They were collected at Waunta Haulover, which is on the Mosquito coast or Atlantic seaboard of Nicaragua. Though none are new, they prove of interest on account of the locality. SILURID^. FELICHTHYS Swainsoii. 1839. Nat. Hist. Fish. Amph. Rep., II, p. 305. Type F [elichthys] filamentosus Swainson, /. c. Based on Silurus bagre Bloch, Ichthyologie, IV, Pt. 11, 1797, p. 19, PI. 365. "Les grand rivieres du Bresil et de I'Amerique septentrionale." (Coll. M. de Burgsdorff Conseiller.) [Not of Linnaeus.] 1. Felichthys marinus (Mitchill). 1815. Silurus marinus Mitchill, Trans. Lit. Philos. Soc. New York, I, p. 433. "New York. Taken June 30th, 1814." Head 3|; depth 4i; D. I, 7; A. viii, 16; P. I, 13; V. i, 5; width of head 1^ in its length; depth of head, at posterior margin of the opercle, If; snout 2f ; eye 5; width of mouth 2; interorbital space If; inter- nasal space 3; base of the dorsal 2|; dorsal spine li; least depth of caudal peduncle 3f ; base of anal If; first branched anal ray If; ventral If. Body rather long, depressed at first, then compressed after the head, and the greatest depth at the origin of the dorsal fin. Caudal peduncle compressed, and its least depth about half its length, measured be- tween the posterior base of the adipose dorsal, and the base of the caudal. Head broad, depressed, more or less flattened below, and rising con- vexly above, especially to the dorsal fin. Upper profile of the head a little convex above the eye in front. Occipital buckler solid, its width in front about equal to its length. Snout broad, a httle rounded when viewed from above, and projecting a Uttle beyond the mandible. Eye lateral, rather large, and its posterior margin about midway in the length of the head, viewed laterally. Mouth broad, rounded, and with 1903.] NATURAL SCIENCES OF PHILADELPHIA. 347 rather thin lips. Teeth minute, and in rather narrow bands in the jaws, and on the vomer and palatines. Tongue thick, broad, rounded, and free in front. Nostrils close together, the anterior circular, directed forward just above the lip, and the posterior elongate with a small flap. Interorbital space convexly elevated, with a median depression giving place to the fontanelle. The fontanelle does not extend beyond the eyes, but runs back to the occipital process as a narrow median groove. Maxillary barbels broad, flattened, and reaching to the anus. Mental barbels short, two in number, and reaching a little beyond the gill-membrane where it crosses the isthmus. Gill-opening extending forward nearly opposite the posterior margin of the eye. Peritoneum pale. Anus about the last third in the space between the origin of the ventral and that of the anal. Skin smooth, the top of the head posteriorly, and the occipital buckler rather rugose. Humeral process smooth. Lateral line with many rather long straight and obliquely vertical branches running out above and below. The lateral line itself is continued well out on the basal portion of the caudal. Origin of the dorsal nearer the tip of the snout than the origin of the anal, the spine straight, with a rugose edge in front becoming barbed above, and giving place above to a long compressed broad fila- ment that reaches near the base of the caudal. Dorsal rays graduated from the first, which is much longer than the spine, to the last, which is a little over a third the length of the spine. Adipose dorsal much higher than its base, and inserted much nearer the base of the caudal than the base of the last dorsal ray. Anal with a sHghtly concave margin, the first developed ray the highest, and the origin much nearer the base of the caudal than the origin of the dorsal. Pectoral with a long compressed spine, rugose along the edge at first, and both edges barbed distally. When depressed the spine reaches three-fourths the distance to the ventral, and its extremity gives place to a long com- pressed filament reaching the origin of the anal. First pectoral ray longer than the spine, and the others graduated to the last which is about a fourth the length of the spine. Ventral broad, rounded, the spine more or less flexible, and reaching the anal fin. Caudal large deeply forked, and the lobes long, rather slender and pointed. Color in alcohol grayish above, white below. Lower side with crowded gray dots. The outer lower surface of the pectoral with 348 i^ROCEEDINGS OF THE ACADEMY OF [May, crowded gray dots. On the outer portion of the anal they are fewer. Ventral with only a few gray dots above. Upper edge of the maxillary barbels with gray dots. Lips whitish. Length 8^ inches. One example. This is evidently the young of marinus, though I have never seen any young examples before. It agrees with adult and rather large examples from the New Jersey coast. H^MULID^. CONODON Cuvier. 1830. Hist. Nat. Poiss., V, p. 116. Type Conodon antillanus Cuvier. /. c. "Jamaique." (Coll. M. Broussonnet.) [=plumien.] 2. Conodon plumieri (Bloch). 1797. Scixma plumieri Bloch, Ichthyologie, III, Pt. 9, p. 57, PL 306- "Antilles." Head 3; depth 3i; D. XII, 12; A. Ill, 7; scales 52 in the lateral line to the base of the caudal; 6 scales between the origin of the dorsal and the lateral line in a vertical series, and 12 between the latter and the origin of the anal ; fourth dorsal spine 1| in the head ; second dorsal ray 2|?; pectoral li; snout S^- in the head, measured from the tip of the upper jaw^; eye 4|; maxillary 2|; interorbital space 5. Back elevated and compressed, the greatest depth at the base of the second dorsal spine. Least depth of the caudal peduncle one and two-thirds in its length. Upper profile of the head a little convex. Snout broad, convex, with the upper jaw projecting, and the upper profile a little convex. Eye high, impinging on the upper profile of the head, and its posterior margin about midway in the length of the same. Preorbital margin undulate. Maxillary reaching opposite the first fifth of the eye, and its distal expanded portion a little less than half. Lips thick and fleshy. Bands of minute teeth in the jaws, with fif large conic canines in the front of each jaw forming an outer series. Tongue broad, round, and free in front. Nostrils large, close together at the front of the eye, circular, and the posterior about half the size of the anterior. Interorbital space flattened. Margin of the preopercle with sparse short serrations, ending in a spine below, and its lower edge with a series of short spines directed forward. . ' Gill-opening extending about opposite the posterior fourth of the eye. Gill-rakers 6+16, pointed, more or less even, and a little shorter 1903.] NATURAL SCIENCES OF PHILADELPHIA. 349 than the filaments, which are a httle less than half the eye. Pseudo- branchiae rather short. Peritoneum pale. Scales of more or less even size, and finely ciliated, ^^ertical fins mostly scaly. Small scales crowded at the humeral region, at the base of the pectoral, and out on the basal portion of the fin. A flap composed of a few small scales at the axil of the pectoral. A few pointed scales at the base of the ventral, but not free. Lateral line nearly concurrent with the dorsal profile, composed of simple tubes, and extending well out on the caudal basally. Origin of the dorsal inserted well behind that of the pectoral, the spines graduated to the fourth which is the longest, the first a little longer than eleventh, and the twelfth about equal to the second. Origin of the soft dorsal about midway between that of the spinous dorsal and the base of the caudal, and the anterior rays the longest. Spinous anal inserted a trifle behind the origin of the soft dorsal, the second spine the longest, much longer than the first or longest ray, and the base of the rayed fin a trifle over half its height. Pectoral pointed, reaching about three-quarters of the distance to the anal. Ventral inserted behind the pectoral, reaching a])out three-quarters of the distance to the anus, and the spine aliout two-thirds the length of the fin. Color in alcohol brown alcove, pale beneath. About eight vertical brown bands or bars along the side of the body, extending till opposite the base of the pectoral. Length Gf^ inches. One example. I do not accept Pcrca nohilis Linnseus^ as the earliest available name for this species, as he states D. |f , A. 3%! MM. Jordan and Feslar con- tend that it is probably intended for this species, though they admit that there is less doubt al^out the name 'plumieri? CIOHLID^. HEROS HeckeL 1841. Zool. Abhand. Annel. Wien. Mus. Nat.. 1841, p. 362. Type Heros severus Heckel, I. c. "Marabitanas im Rio-negro." (Coll. M. Natterer.) 3. Heros urophthalmus Giinther. 1862. Heros uwphthaJmus Giinther, Cat. Fish. Brit. Mus., IV, p. 291. "Lake Peten." [Guatemala.] (Coll. MM. Salvin and Godman.) Head 2^; depth 2^; D. XVI, 11; A. VI, 9; scales 30 in a lateral 1 Syst. Nat., Ed. X, I, 1758, p. 291. 2 Rep. U. S. Fish. Comm., XVII. 1893, p. 488. 350 PROCEEDINGS OF THE ACADEMY OF [May, series to the base of the caudal; 5 scales in an oblique series back from the origin of the spinous dorsal to the lateral line, and 11 between the latter and the origin of the anal; fifth dorsal ray ly in the head; sixth anal spine 2^; fifth dorsal ray If; caudal 1^; pectoral H; ventral If; least depth of caudal peduncle 2f ; snout 2f in the head, measured from thetipof the upper jaw; eye 4|; maxillary 3yV; interorbital space 3^ ; width of preorbital 5. One example, 5j\ inches in length. It is more slender than the examples examined by MM. Evermann and Goldsborough,^ and the black caudal spot at the bases of the upper caudal rays is very distinct. There is also a certain amount of dusky about the bases of the vertical fins, and the outer anterior portion of the ventral. TETRAODONTIDuE. 4. Cheilichthys testudimeus (Linnseus). Four small examples, the largest 3f inches long. They agree with examples from Porto Rico. There are a number of rather large dark spots on the flank, but not so numerous as in the adult. SOLEIDuiE. BAIOSTOMA Bean. 1882. In Goode and Bean, Proc. U. S. Nat. Mus., V, p. 415. Type Baiostoma hranchialis Bean, I. c. "Appalachicola Bay, South Florida." (Coll. M. Silas Stearns. Nos. 26,605, 30,463, U. S. Nat. Mus.) This group is given generic rank on account of the presence of small pectoral fins, developed at least on the right side. In Achirvs they are wholly wanting. Grammichthys Kaup is hardly available as his type, PIcuronectes lineatus Linnseus, appears to be of the Ed. XII, Syst. Ahit. 5. Biastoma lineata (Linnseus). 1758. [Pleuronedes] lineatus Linnseus, Syst. Nat., Ed. X, p. 268. "Amer- ica." Based on Pleuronedes 1. Fuscus suhrotuncL, etc. Browne, Civ. Nat. Hist. Jam., Ill, 1756, p. 445. [Jamaica.] D. 53; A. 40; scales 76 in a lateral series between the gill-opening and the base of the caudal. Pectoral well developed on the right side with five rays. One example, 2f inches in length. 3 Bull. U. S. Fish Comm., 1902, p. 157. 1903. NATURAL SCIENCES OF PHILADELPHIA. 351 May 19. The President, Samuel G. Dixon, M.D., in the Chair. Twenty-one persons present. Regeneration of the Body of a Starfish. — Miss Sarah P. Monks, in a note read to the meeting, stated that it had been known for many years that starfish can renew rays that have been removed. " It has been stated that in certain starfishes an arm itself can produce a new star- fish — Haeckel, Sarasin, von Martens and Saris — but this has been denied by other observers. "^ It is beheved by some that a portion of the disk must remain if there is any regeneration, and the breakage plane is near the body, or disk, in cases of renewal of arms. In studying regeneration of Phataria {Linckia) fascialis she had cut arms at different distances from the disk, and a number of the single rays produced new bodies. The free ray made a new body and the rest of the starfish produced a new ray, and there was very little differ- ence in the rate of growth of each, and no definite place for breaking. In the photograph of a six-rayed Phataria, the cut ray attached to the body shows a small ray sprouting, while the free ray shows four new rays. This was cut July, 1902, and the photograph taken Febru- ary, 1903. The manner of growth is as follows: The cut edges heal and draw- down toward the oral side of the starfish, then small knobs appear at the end which grow into rays in which the ambulacral furrow soon appears, wdth the small mouth in the center of the rays. She had collected specimens at San Pedro, Cal., showing all stages of growth of the single arm, from the recently broken arm to those like the photograph on through all sizes of growing rays. The following have been accepted by the Publication Committee and ordered to be printed : ^Regeneration, Morgan, p. 102. 352 PROCEEDINGS OF THE ACADEMY OF [May, THE FORM AND STRUCTURE OF THE MYCODOMATIA OF MYRICA CERIFERA L. BY JOHN W. HARSHBERGER, PH.D. Within the last decade or two considerable progress has been made in our knowledge of the enlargements, galls, tubercles and coralline outgrowths on the roots of the higher chlorophyll-bearing plants. Some of them are due to insects, others are due to a perversion of the physiologic activities of the plants on which they are found, while others are attrilDutal^le to the stimuli occasioned by bacteria, slime moulds and higher fungi. Our information concerning the tubercles on the roots of the Leguminosse is reasonably complete, thanks to the energies of Hellriegel, Willfarth, Winogradsky and others. Magnus^ has summed up our knowledge of the growths joroduced by subterra- nean fungi in a recent paper. He describes systematically the fungi known to live as subterranean parasites, but barely mentions those forms of enlargement called mycodomatia. Mycodomatia were known to the botanists of a century ago. Meyen^ looked upon them as para- sites having a habit in this respect similar to plants of the natural orders Balanophoracese and Orobanchacese. Schacht,^ who was the first to give a fairly satisfactory account of their external appearance, regarded them as normal growths upon roots, but later he considered them as abnormal. Jager* considered them as due to insects. Wo- ronin,-^ in a paper published in the Memoirs of the Academy of Sciences of St. Petersburg, believed that the coral-like swellings on the roots of the black alder were due to a fungus closely related to one described by Nageli inhabiting the roots of various species of Iris and called by him ^Magnus, P., "Unsere Kenntniss unterirdisch lebender streng parasitischer Pilze und die biologische Bedeutung eines solchen unterirdischen Parasitismus," Abhandhingen des hotanischen Vereins der Provinz Brandenburg, XLIV (1902), pp. 147-156. - Meyen, "Ueber das Hervorwachsen parasitischer Gewiichse, etc.," Flora, 1829, .S. 49. ^ ScHACHT, "Die Pflanzenphvsiologie und Herr Dr. G. Walpers in Berlin," Flora, 1853, pp. 1-13; also "Der Baum," 18(50, S. 172-174. ^ Jager, "Ueber eine Krankhafte Veriinderung der Bliithen Organe der Weintraube," Flora, S. 49. * WoRONiN, "Ueber die bei der Schwarzerle (Alnus Glutinosa) und der gewohn- lichen Garten-Lupine (Lupinus mutabilis) auftretenden Wurzelausschwellungen," Memoires de V Academie ImperiaJe des Sciences de St. Pctersbourg, VII Serie, Tome X, No. 6, 1866. 1903.] NATURAL SCIEXCES OF PHILADELPHIA. 353 Schinzia celhdicola. ^lagnus denied that the similarity was sufficient to class the two fungi together. Some work done by Gravis" led Woronin to make a more detailed study of young galls upon Alnus. Woronin was so much impressed with the resemblance to his Plasmo- diophora brassicce, that he communicated to Gravis his belief that two organisms were present, one a slime mould and the other a fungus. Since these preliminary observations similar galls have been found on Alnus incana, Alnus serridata, Abius undulata, Ceanothus ameri- canus, Myrica gale, Hippophce rhamnoides, and on species of Elceagnus and Shepherdia, and last summer on Myrica cerifera by the writer. The results of his study of the mycodomatia on the roots of the common waxberry form the material for this paper. Tabulated the names of the hosts on which mycodomatia occur and the names of the fungi producing them is as follows : Alnus glutinosa Frankia alni (Moller) Atkinson. '' incana -... " " serrulata " '' undulata " Ceanothus americanus Frankia ceanothi Atkinson. Myrica gale Frankia brunchorstii Moller. " cerifera " Elceagnus — - Frankia cdni (Moller) Atkinson. HippophcB rhamnoides ..— " " " " Shepherdia '' A history of the synonymy is somewhat as follows: Woronin first considered the parasite to belong to the genus Schinzia of Nageli. Magnus showed the untenableness of this position. Woronin, after more detailed study, referred the parasite to the slime mould genus Plas- modiophora, and in this view he was supported by H. Moller,^ who called it Plasmodiophora alni. Brunchorst,^ correctly interpreting the nature of the parasite to be a true filamentous fungus, estabUshecl the genus Frankia and named it Frankia subtilis, ignoring the oldest specific name alni. Morller in 1890, by a study of fresh material instead of alcoholic, set aside his former position and corroborated the observa- tions of Brunchorst that the organism was a filamentous fungus. He * Gravis, "Observations anatomiques sur les excroissances des racines de I'aune," Bulletin de la Societe royale de Botanique de Belgique, Tome XVIII, partie I, pp. 50-60. "Moller, H., "Plasmodiophora alni," Berichte der deutschen botanischen Gesellschaft, Bd. Ill, 1S85, pp. 102-105. *^ Brunchorst, "Ueber die Knollchen an den Wurzeln von Alnus und den Elgeagnaceen," Botanisches Centralblatt, XXIV, p. 222, 1885. 23 354 PROCEEDINGS OF THE ACADEMY OF [^lay, also described the parasitic growths on another plant, viz., Myrica gale, and named it as a new species, Frankia hruncJiorstii.^ Atkinson*" gave a complete and useful summary of the literature when he pub- lished his paper in 1892. He described a new species of Frankia pro- ducing galls on the roots of the New Jersey tea, Ceanothus americanus, which he called Frankia ceanothi. The mycodomatia on Elceagnus, Hippophm and Shepherdia are considered to be due to a parasite iden- tical with Frankia aim occurring on the genas Alnus. We have, therefore, three named species of this genus, viz., Frankia alni (Moller) Atkinson occurring on four distinct genera of phanerogamous plants, Frankia ceanothi Atkinson on the single genus Ceanothus, and Frankia briinchorstii Moller on the genus Myrica. The discovery of mycodomatia on a new host, viz., Myrica cerifera, has led the writer to make a careful study of the galls and the fungus that produces them. The mycodomatia were discovered on waxberry bushes growing on the slopes of sand dunes at Sea Side Park, N. J., where they were laid bare l^y the blowing away of the sand from about the plants. Originally the bushes grew out of a low dune, and as the .sand drifted in about their stems adventitious roots were formed upon which grew the fungous galls, or the mycodomatia. As will be shown subsequently, the fungus is perennial and the growth of the m3'codo- matia is an extremely slow one. This slow growth argues for the stabiUty of the dune on which the bushes grew, for if the sand had been constantly shifting the roots could not have been buried sufficiently long to permit of the growth of the fungous galls to the size that they had reached when they were uncovered. We have, therefore, a means of measuring the length of life of certain sand dunes. It seems to the writer that the discovery of these mycodomatia on the adventitious roots of the waxberry growing in the pure sand of the New Jersey dunes throws light upon the question of the importance of these mycodomatia to the host plant. It has been shown by the experimental work of a number of observers that leguminous plants will grow in pure sand after the period of star- vation is passed, provided such sand be microbe-seeded, i.e., provided the right kind of tubercle bacterium is present in the sand. Subse- quent chemical analysis of plants thus grown has shown that nitrogen over and above the amount present in the seed is found in such legu- ^ Moller, "Beitrag zur Kenntniss der Frankia subtilis Brunchorst," Bericht der deutschen hotanischen Gesellschaft, VIII, 1S90, pp. 215-224. 1" Atkinson, George F., "The Genus Frankia in the United States," Bulletin of the Torrey Botanical Club, XIX (1S92), pp. 171-177. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 355 minous plants, and the inference is that the tubercle-producing bacteria have been instrumental in the production of various nitrogenous compounds derived directly from the nitrogen of the atmosphere. It is customary in the poor sandy soil of northern Germany, near Berlin, to grow a great variety of leguminous plants, prominently the yellow luj^ine. The yellow lupine is plowed under and enriches the sandy soil by the decomposition of the nitrogenous substances present in the roots, stems and leaves. The fact that such plants thrive in such poor sandy soil is explained by the activity of the bacterial symbiont. The writer has observed the beach pea, Lathyrus maritimus, growing on the crest of the sea dunes of the New Jersey coast. That the soil is enriched by the growth of this plant is evidenced by the more luxuri- ant development and darker green color of the marram grass, Ammo- phila arenaria, which grows associated with the beach pea on the dunes. Thus the writer previously argued, taking for granted that Frank's assumption was correct, that mycodomatia have a similar function to the leguminous tubercles. But a more careful study has led him to believe that too much has been taken for granted with reference to the function of mycodomatia. The abundance of the mycelia in the mycodomatia surprised him and led him to question the validity of the position taken by some of the earher observers that the mycodomatia act in the same way as the leguminous tubercles. It is probable that Frankia brunchorstii is more in the nature of an endotrophic mycorhiza, to be placed inter- mediate between the ectotrophic mycorhiza found on the Indian pipe Monotropa, the short roots of which resemble closely in external appear- ance mycodomatia, and the typical endotrophic mycorhiza found in Thismia and certain other plants, where a definite relationship is established between the nucleus of the host and the fungous hyphse. No such nuclear control of the growth of the fungus Frankia is observed in the mycodomatia of Myrica cerifera. The action of the mycelium of Frankia is much more severe, and in fact its growth sug- gests a true antagonistic symbiosis, for the host cells finally suffer the loss of their protoplasmic contents and collapse, leaving the fungus in possession of the older portions of the mycodomatia. Whether the waxberry derives any benefit from the association of Frankia with its roots can be determined only b}^ careful physiologic experimentation. However, if we have here a true instance of para- sitism, the struggle between host and fungus is long drawn out, and no material damage is done to the host as long as the fungus con- fines its attack to the secondary roots of the waxberry. If this view 356 PROCEEDINGS OF THE ACADEMY OF [May, is correct, then the galls on the roots of Myrica are mycocecidia and not mycodomatia, applying these terms as first suggested by Frank. The mycodomatia (PI. XVI) examined by the writer all grew upon the short adventitious roots formed when the stems of the waxberry bushes were covered by the blowing of the sand around them. On some smaller secondary roots the galls simply dichotomize, but later, by the increase in number of these forking fiber-Hke swelHngs, they become aggregated together into nests or clumps about the size of a walnut (PI. XVI). The dichotomous fibers that compose the myco- domatia are of a rich umber-brown color. They grow in length by small increments and repeatedly branch in a forking manner. On a small stem examined, the fibers developed on the adventitious roots surround the dry remains of the underground rhizomes of the marram grass, Ammophila arenaria. The dead leaves and wiry stem of this grass are mixed with the fiber-like galls by the repeated branching of the galls among this material. The tips of the brown fibers that to- gether form a fungovis household, or mycodomatium, are in the dried specimens blunt and rounded. Their appearance seems to indicate that, when fresh, they were of a lighter color and softer in consistency than the older part of the swelhng. The fighter color of the tip prob- ably indicates the growth of the year. A measurement of several such apices shows that the growth is extremely slow, rarely exceeding a millimeter or two in a single season. Some of the branches of the mycodomatia measure twelve and fourteen millimeters in length. If the yearly increment is one millimeter, such branches are twelve and fourteen years old. If the annual growth is two millimeters, six or seven years represent the age of some of the branches. A conserva- tive estimate of the age of the mycodomatia that have reached the size of walnuts is ten to fifteen years. If the growth in the length of the branches of the mycodomatia is greater than this, then this esti- mate is too high. One waxberry stem thirty-two millimeters in diam- eter with several mycodomatia on its secondary roots shows twenty- two annual rings of wood, and twenty years would be the outside limit of the age of mycodomatia growing on such stems. When dried tlie branches of the mycodomatia become extremely brittle, and the speci- mens kept for the botanical museum suffered severely in being carried from the seashore to the botanical laboratory. The microscopic structure of the galls is of interest because few of the earlier observers seem to have determined satisfactorily the exact character of the parasite. Thus Woronin" considered the parasite to " WoRONiN, loc. cit. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 357 be a fungus similar to Nageli's Schinzia. A paper by Gravis led him to modify his views by ascribing the galls to the combined action of a myxomycete similar in appearance to his Plasmodiophora hrassicce and a fungous mycelium. Moller^^ claimed that the galls were due to a slime mould. Warming^^ attributed the formation of the mycodo- matia to a shme mould allied to the genus Plasmodiophora. Brun- chorst," by his excellent observations, set the matter straight by attribu- ting the galls to a filamentous fungus and estabhshed the genus Frankia for it. Woronin, Frank/^ Sorauer^^ represented several so-called spo- rangia in the cells of the several hosts studied attached to single threads of the mycelium. Brunchorst attempted to prove the fallacy of the observations of these workers by showing that by an optical illusion the sporangia which appear attached in reality lie over the fungous hyphse. Atkinson^^ figures and describes the myceUum and sporangia of a filamentous fungus which he called Frankia ceanothi, because the parasite lived in the roots of the New Jersey tea, Ceanothus americanus. With this contradictory evidence a more careful micro- scopic examination of the mycodomatia is necessary. Sections were made of the branches of mycodomatia by first boiUng the dried specimens and then treating them with thirty-five per cent, alcohol to remove part of the air. Transverse and longitudinal sec- tions were made of the dichotomously branched root-like galls. The general microscopic structure of one of these mycodomatial sweUings resembles that of a root (PI. XVII, fig. 1). The center of the section is occupied by the cylinder of wood or xylem, which, however, lacks the larger open elements of the wood of a normal root. The tracheids, irregular in shape and much reduced in size, are compacted together and the medullary rays are displaced out of their true radial position, taking a somewhat sinuous instead of a straight course (PI. XVII, fig. 1). Both in the normal and in the fungous-infested tissues, the medullary ray cells have contents of a rich brown color. External to the wood comes the cambium, theoretically of a single layer of cells, and outside of this the soft bast which consists of rounded cells. In such roots, where the elements have shifted normally from a radial jDosition, the cortex and soft bast are confluent, both in the normal and in the fun- ^- MOLLER, loc. Cit. 13 Warming, " Wurzelknollchen bei den Elseagnen," Just's Botanischer Jahres- bericht, 1876, IV a, p. 439. 1* Brunchorst, "Ueber die Knollchen an den Wurzeln von Alnus und den Elseagnaceen," Botanisches Centralblatt, XXIV, 1885, p. 222. '^ Frank, "Krankheiten der Pflanzen," p. 647. / ^® SoRATJER, "Pflanzenkrankheiten." 1' Atkinson, loc. cit. 358 PROCEEDINGS OF THE ACADEMY OF [May, goiis material. In the older normal roots, the cortex is dehmited by a discontinuous layer of hard bast patches, a few elements of which are occasionally met with in the galls. External to the rather abundant brown cortex region of normal and fungous inhabited roots is a phello- genetic layer, succeeded at the periphery by the young and old cork cells. Where branches arise, a section at such places shows the obhquely cut xylem pushing out surrounded by the cortex cells. The young hght-colored cork cells at such places become confluent \vdth the similarly colored wood cells, so that it is difficult to distinguish between the elements composing these two distinct kinds of tissue. Another marked feature in both the normal and parasitized roots is the plugged tracheids with a yellowish or brown gummy material,^* whether in the nature of modified tyloses the writer was unable to determine. The most highly modified portions of the roots of Myrica cerifera, when parasitized by Frankia brunchorstii, are the woody cylinder, the soft bast and the cortex. The finest fungous mycelium is found in the cortex of the younger roots and growing into the medullary ray cells. It consists of fine unicellular hyphse and can be made out with the greatest difficulty by a No. 3 Leitz objective. With a No. 7 Leitz objective, the finer hyphae become defined as cobweb-like threads stretching across the large, lacunar, intercellular spaces which have been formed between the rounded cortex cells near the apical portion of the swelling (PL XVII, fig. 66). Sometimes the hyphae stretch straight across these inter- cellular spaces, but more often they take a sinuous course and form a complex where several branches cross each other (PI. XVII, fig. 3). These finer hyphse are formed as branches from thicker brown hyphse to be described later. The course of the hyphse, as revealed in a longitudinal section of the apical portion of the mycodomatial branches, is in general from cell to cell. The hyphse may pass from one side of the cell to the other, pass- ing out again through the cell wall, or the hyphse may make a loop, a half turn or branch by the formation of short branches (PI. XVII, fig. 6). These short branches are found imbedded in the protoplasm of the cortex cells and may be looked upon in some sense as haustoria. Sometimes several hyphse run into one host cell, and in such cases the branches form a mesh. The hyphse also grow intercellularly. In several transverse sections studied, three hyphae parallel to each other pierce the same cell wall and run through the protoplasm of the cell ^^ Harshberger, "Two Fungous Diseases of the White Cedar," Proc. Acad. Nat. Sci. of Phiki., 1902. p. 461. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 359 thus entered. The course of such hyphse from cell to cell is made clearer by the contraction of the protoplasm from the cell wall. This condition has been produced by drying and the protoplasm has been plasmolyzed. Two hyphse enter the protoplasm of a cortex cell, converge near the center of the cell, and then run to the opposite cell wall, where just before passing through it they diverge from each other. Another hyphse enters a cell, and in the center of it forks to form a Y (PI. XVII, fig. 6&). Still other hypha? in transverse section grow through the triangular intercellular spaces, forming perfect complexes. In the older sections, hyphse are seen of a larger size and with browner walls than the finer hyphse above described (PI. XVII, figs. 2, 4). These seem to take a general longitudinal or oblique course through the cortex, because in several transsections studied these hyphse exist as rings lying in the cells, having been cut across by the razor. Large brown hyphse are seen in the lacunar intercellular spaces of sections made at the base of mycodomatial swelhngs (PI, XVII, figs. 2 and 4). These hyphse are the main trunks of those that pierce the cells and grow into the protoplasm, for they produce smaller branches which assume the colorless aspect of the finer hyphse already described (PI. XVII, fig. 66). It is, therefore, clear that the apical portions of the mycodomatia have fine hyphse with a few thicker strands, while sections cut from the older and basal portions of the swellings have large, brown, thick-walled, unicellular hyphse which run longitudinally and obhquely. The larger hypha probably form the older and peren- nating mycelium which, during the hfe of the metamorphosed second- ary roots, seem to provide new and finer hyphse to the apical portion of the branches of the mycodomatium. The larger unicellular hyphse, which can be followed across the large irregular lacunar spaces formed by the rupture of the cortex in drying, enter cortex cells where they branch by the formation of short rounded sickle-shaped branches (PI. XVII, fig. 3). Several of these curved branches may be formed from a single hypha. These may be looked upon in the nature of haustoria. Sometimes these branches, especially near the apical portion of the mycodomatia, become extremely fine, and then they may grow between the starch grains imbedded in the protoplasm, forming a meshed structure to be referred to later (PI. XVII, fig. 6a). The larger number of these brown thick-walled hyphse are found in the medio- and endocortex, and they almost fill both the cells of these regions and the intercellular spaces. The cortex cells are no longer living in these reigons, but by the growth of the parasitic hyphse they have been destroyed as living cells. However, at the 360 PROCEEDINGS OF THE ACADEMY OF [May, apex of the mycoclomatial growths the cortex ceUs are still ahve and by slow growth add to the length of the several branches, forming a mycodomatium. This stage of the fungous development is sometimes found on secondary roots which still show a radial structure with a well-defined endodermis. In such roots the fungus is found especially well marked in the medio- and endocortex and to a limited extent in the exocortex. One section showed an appearance suggestive of sporangia as figured in Woronin's paper of 1866, cited above. The writer refers to cer- tain cortex cells which have contents not only reticulate with clear rounded areas enmeshed by the yellowish reticulum, but also suggestive of a lattice-work of protoplasm (PI. XVII, fig. 6a), Hyphae are con- nected not only with the protoplasmic reticulum, but also with the open basket-like protoplasm, so as to suggest that the protoplasmic reticulum owes its origin to the mycelium. A careful study, however, of the relationship of fungus and host cells shows that the reticulum owes its genesis to imbedded starch grains which have been partially dissolved away by the treatment of the sections in mounting, and that hyphse have sent in short branches between the starch grains and hence into the meshes of the protoplasmic reticulum (PI. XVII, fig. 6a). This may have suggested to Woronin the sporangia (zooconidia) which he figures in a grape-like bunch in the cell, each sporangimn (zooconidium) connected by a hypha. Or this reticulate structure may have suggested to Moller^^ a plasmodium of a myxomycete like Plasmodiophora dividing up into a number of spores. The writer believes that in suggesting this he has reconciled the earlier opposing views. Woronin is probably right in describing the sporangia (zooconidia) of Frankia, because the mycelium and its man- ner of growth suggests a relationship to the genus Pyihium, and the writer would place, therefore, tentatively, the hyphomycetous genus Frankia among the Oomycetes, close to the genera Pythium and Perono- spora. This view is strengthened if the lattice-like reticulum mentioned above (PI. XVII, fig. 6a) is compared with a figure (fig. 28) given by Tubeuf on p. 139 of his text-book. This figure illustrates the growth of the fungous Phytophthora in the tissues of the leaf of the beech, and the same kind of reticulum is shown. In the absence of oogonia and zooconidia, however, in the mycodo- matia of Myrica cerifera, the suggested relationships of the fungus studied by the writer to the Oomycetes cannot be insisted upon. The ^' MoLLER, "Beitrag zur Kenntniss der Frankia subtilis Brunchorst," Berichte ^er deutschen botanischen Gesellschaft, VIII, 1890, p. 222. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 361 unicellular hyphse, the method of growth of the haustoria and the appearance of the sporangia (zooconidia) figured by Woronin are all strongly suggestive of such a kinsliip. Bibliography. 1829. Meyen. Ueber das Hervorwaclisen parasitischer Gewachse. Flora, 1829, S. 49. 1853. ScHACHT. Die Pflanzen-Physiologie und Herr G. Walpers in Berlin. Flora, 1853, S. 10. I860. Schacht. Der Bamn, S. 172-174. 1866. Woronin, M. Ueber die bei der Scliwarzerle (Alnus Glutinosa) und der gewohnlichen Garten-Lupine (Lupinus mutabilis) auftretenden Wurzel- anschwellungen. Memoires de VAcademie Imperiale de St. Pitersbourg, VII Serie, Tome X, No. 6. 1866. 1876. Warming, E. Smaa biologiske ogmorfologiske Bidrag. Botanisk Tidsskrift, 3 dei Rakke, Bd. 1, 1876, p. 84. Reference, Wurzel-Ivnollchen bei den Elseagneen, Just's Jahresbericht, 1876, p. 439. 1879. Magnus. Sitzungsberichte des bot. Ver. d. Prov. Brandenburg, XXI, June 27, 1879. 1880. Frank, B. Die Krankheiten der Pflanzen. Breslau, 18S0, p. 648. 1885. Brunchorst. Ueber die Knollchen an den Wurzeln von Alnus und den Elseagnaceen. Botanisches Centralblatt, XXIV (1885), p 222. 1885. Gravis, A. Note sur les excroissances des racines de I'aune. Compte rendu de la seance mensuelle du Jan. 10, 1880, de la Societe royale de Botanique Belgique. See Berichte der deutschen botanischen Gesellschaft, Bd. Ill, 1885, p. 177. ■ . SoRAUER. Pflanzenkrankheiten. 2 Aufl., I, p. 747. 1885. MoLLER, H. Plasmodiophora Alni. Berichte der deutschen botanischen Gesellschaft, III, 1885, pp. 102-105. 1886. Brunchorst. Ueber einige Wurzelanschwellungen, besonders die- jenigen von Alnus und den Ela?agnaceen. Unters. bot. Inst. Tubingen, Zweiter Band, pp. 151-177. Also Botanisches Centralblatt, XXVII, p. 109. 1887. Frank, B. Sind die Wurzelanschwellungen der Erlen und Elseaganceen PilzgaUen? Berichte der deutschen botanischen Gesellschaft, Bd. V, Heft 2, 1SS7, p. 50. Ref. in Botanisches Centralblatt, Bd. XXV, p. 366. 1890. MoLLER. Beitrag zur Kenntniss der Frankia sub tills Brunchorst. Berichte der deutschen botanischen Gesellschaft, Bd. VII, 1890, pp. 215-224. 1891. Atkinson, George F. The Tubercles on the Roots of Ceanothus. Bo- tanical Gazette, XVI, p. 262. t 1892. Atkinson, George F. The Genus Frankia in the United States. Bul- letin of the Torreij Botanical Club, XIX (1892), pp. 171-177. 1895. TuBEur, K. F. vox. Pflanzenkrankheiten durch Kryptogame Parasiten verursacht. Section on Die Mj-codomaticn der Erlen, Eteagnaceen und My- ricaceen veranlasst durch Frankia-Arten, p. 117. 1900. WooDHEAD, T. W. On the St.-ucture of the Root Nodules of Alnus glutinosa. Rep. British Association Advancement Science, 1900, p. 931. 362 PROCEEDINGS OF THE ACADEMY OF [May, EXPLAXATIOX OF PlATES XVI AND XVII. Plate XVI. — Stem of the waxberry, Myrica cerifera, with secondary roots upon which are formed the mycodomatia, or mycocecidia. Photograph by Mr. W. H. Walmsley. Plate XVII, fig. 1. — Transverse section of a branch of a mycodomatia made below the middle showing large lacunar areas, magnified 25 diameters. Fig. 2. — Cortex of mycodomatium showing large thick-walled unicellular hyphte. Fig. 3. — Meshed structure of finer hyphfe from mediocortex. Fig. 4. — Coarser hyphte at the base of a mycodomatial branch. Fig. 5. — Older normal secondary root of the waxberry showing central woody cylinder, etc. Fig. 6. — Finer hj^jhte with one cell at a showing meshed structure suggestive of a collection of zooconidia. At h finer hyphaj are sho^vn. Photo- graphs by Mr. W. H. Walmsley. 1903.1 NATURAL SCIENCES OF PHILADELPHIA. 363 ON THE MORPHOLOGY OF THE ROTATORIAN FAMILY FLOSCULARIID^. BY THOMAS H. MONTGOMERY, JR.^ The thoughts which group themselves about the theories of larval homologies, and in particular that one which regards the trochophore stage as the recurrence or at least parallelism of an ancestral phyletic form, comprise one of the many inducements to investigate the anat- omy of the Rotatoria. For it is generally maintained by those workers who uphold the trochophore theory in its logical sequences, that the adult Rotatorian and the trochophore larva show close similarities. One line of the approach, then, to test the correctness of the ideas of such larval homologies, is to extend our knowledge of the structure of the Rotatoria. A huge literature has grown up around this group of animals, but with the exception of a very few detailed monographs the writers have contented themselves w^th the description of the exter- nal form. This is the more remarkable since the Rotatoria lend them- selves better than most forms, thanks to their great transparency, to the study of fine details of structure in the Hving animal. Those who have not occupied themselves especially with the anatomy of this group make the common assumption that all Rotatoria show essentially the same type of structure. But this is erroneous for two reasons : first, because in Nature there do not occur types of structure, but gradations of structure; and second, because the Rotatoria evince very great differences among themselves. For the bearing of the group as a whole upon the views of larval homologies it is first necessary to examine the various Rotatorian groups comparatively, in order to determine which of them is the most generahzed or primitive; then this group should be compared with the various kinds of trochophore larvae. This is a problem to be attacked from more sides than the purely morphological, for at once appears the striking fact that the Rotatoria are for the most part fresh-water forms, while the trocho- phore larvae are characteristic exclusively of marine groups. Of the living Rotatoria, are the most primitive forms found in the sea or in fresh water? And of them, are the pelagic forms more primitive than 1 From the Zoological Laboratory of the University of Pennsylvania, Phila- delphia. 364 PROCEEDINGS OF THE ACADEMY OF [-Mav, the creeping and attached forms? Here again we are met with the fact that a knowledge of distribution, and its factors, constitutes a very important ally in the study of phylogeny. The present paper is offered as a contribution to the morphology of the Flosculariidse alone. The only thorough account so far of the anatomy of any species of this family is that of Gast on Apsilus lenti- formis. The occurrence of Apsilits, Stephanoceros and four species of Floscularia during the past winter and spring in a pond in the garden attached to our zoological laboratory, enal^led me to study all these forms at the same time, and so to make the desired comparison of them. But until I have had opportunity to examine for myself other families of the group, it would l)e premature to express any opinions upon the phyletic value of the Rotatoria as a whole. In regard to the methods of stud\^ the examination of the living animals under slight pressure of the cover-glass has proved the most important. Fixation with hot corrosive sublimate or with Flemming's fluid, and staining with hsematoxylin or carmines were of value in rendering nuclei more distinct, but even in life all the nuclei of the body tissues can be seen. The preparations may be mounted in l^alsam with no shrinkage by passing the objects from the absolute alcohol through graduated mixtures of cedar oil vdth alcohol. Sections were made of Apsilus alone. The complete literature has been cited for Apsilus and Stephanoceros, but for Floscularia only such contributions as concern the internal anatomy. APSILUS, Meczn. Historical. Leidy (1857) described very briefly Dictyophora nov. gen. vorax nov. sp., from the vicinity of Philadelphia. Mecznikow (1866) described Apsilus nov. gen. lentiformis nov. sp., from Giessen; he noted the com- plete absence of ciliary wTeaths, form of the corona, the lateral antennae, intestinal tract, musculature, nephridia, and mistook for the brain a large muscle; he described the male also. Then Leidy (1882) figured his species, and mentioned that it differs from the species of Mecznikow in the lack of lateral antennae and of a ganglion. Forbes (1882) gave a crude figure of and briefly described Cupelopagis nov. gen. bucinedax nov. spec, from Illinois, describing the alimentary tract and the exter- nal form. Foulke (1884) described as a new species Apsilus bipera, from Philadelphia ; the corona was described as a membranous hood or net, two stomachs (the proventriculus erroneousl}' regarded as a stomach), lateral antennae, musculature; she pointed out that the 1903.] NATURAL SCIENCES OF PHILADELPHIA. 365 name Didyophora is preoccupied, and proposed to recognize" as three distinct species Apsilus vorax (Leidy), A. bipera Foiilke and A. biicine- dax (Forbes). In rejoinder to this paper Leidy, in the same 3^ear, pro- posed to unite these under the name Apsilus vorax, and found at last the lateral antennae. Hudson and Gosse (1886) placed vorax and hucin- edax as synonyms of lentiformis, and hipera as a distinct species ; they relegated this genus to the FlosculariidcB. Stokes (1896) gave a good description of bucinedax, from Trenton, N. J.; he described the lateral antennae, oesophageal tube, long immobile setae arranged in tufts upon elevations of the inner surface of the coronal cup, slender and numerous coronal muscles, an ovary extending across the whole width of the body. He also figures the corona of bipera from the same locahty, and considers the four described species as all distinct. Lund (1899) held that the genus should be removed from the Flosculariidce, and placed in a separate family near the Asplanchnidce. Jennings (1900) figured the young of a species of Apsilus. Finally Gast (1900) has given a very detailed and accurate account of the anatomy and histology of a species he calls vorax Leidy (holding lentiformis to be synonymous), and regards bipera and bucinedax to be distinct from it. Thus the European species, lentiformis, has been well described by Mecznikow and Gast, but of the three described American species the accounts, with the exception of that of Stokes of bucinedax, are so meager that the status of these is very perplexing. Foulke was right in uniting all these in the one genus Apsilus. A. bucinedax (Forbes) is well marked from all the others by the shape and great size of its germariurii. Now the species which I shall describe is from the same locality as the forms described by Leidy and Foulke, and agrees with both of these in all essential particulars (these authors had overlooked the germaria, nephridia and foot, all the sense-organs except the lateral antennae, and the oesophageal tube). Foulke considered her species to differ from Leidy's in possessing lateral antennae (but these had been overlooked by Leidy), in cihation of the corona (probably, as Gast has remarked, flame cells were mistaken for such ciliation), in more strongly developed muscular system, and in the external form. I am inclined to conclude that these differences are only apparent, due to the meagerness of Leidy's descriptions. Stokes stated that bipera differed from vorax in that the ventral margin of the corona is projected forward; but in the form described by me this margin shows consid- erable variations in form. Hence bipera (Foulke) had best be consid- ered a synonym of vorax (Leidy); and the three clearly recognizable species may be distinguished as follows: 366 PROCEEDINGS OF THE ACADEMY OF [-^iay, (a) — Gcnnarium very large, extending across the whole width of the trunk, oesophageal tube present, immobile setae within the coronal cup, bucinedax (Forbes). (6) — Germariura small, rounded, no setae within the corona. (1) — -(Esophageal tube present, dorsal cuticula of trunk with trans- verse thickened ridges, vorax (Leidy). (2) — CEsophageal tube absent, dorsal cuticula without such ridges, lentijormis Mecznikow. In the description which follows I shall refer for comparison mainly to the description of Gast for lentijormis, since his account is by far the most thorough. Anatomical, The external form is fairly well known. The trunk is somewhat wider than high, rounded posteriorly, while anteriorly it is continued as a large corona (Riissel, Mecznikow; cup, Leidy; net, Foulke; Muiid- trichter, Gast). When fully extended the form is as shown in the figures (PI. XVIII, 1-3), the corona very large and its aperture usually in the horizontal plane but sometimes oblique. The posterior wall of this aperture (mouth) may be rounded or notched, or may project forward as a convex lobe. On the ventral surface of the trunk is the rudimentary foot (Chitinring, Mecznikow; disk or sucker, Leidy; Fuss, Gast) ; and behind that, also ventral, the cloacal aperture. The lat- eral antennae are clearly visible on the sides. The whole animal is so beautifully transparent that the greater part of the following descrip- tion has been based upon a study of the living animal. Hypodermis and cuticula. — The hypodermis is a very thin layer, with its flattened nuclei far apart; at the margins of the mouth it is only slightly thickened. The cuticula is thin, colorless, and (especially on the dorsal surface) covered with minute tubercles which may be ovoid or stellate (as Gast found). On the antero-dorsal surface of the trunk there are broad but narrow thickenings of the cuticula (Pl.XVUI, X, fig. 1), the number and exact arrangement of which is subject to considerable variation; each of these ridges bears tubercles. The cuticula and hypodermis of the body wall are continued inward as the lining of the corona; there the minute tubercles are replaced by minute conical projections, borne mainly upon longitudinal ridges of the cuti- cula. The whole cuticula is very flexible, and when the animal is strongly contracted and the corona rolled inward it is thrown into many folds; radial folds are also found around the cloacal aperture. No cilia are found anywhere upon the surface of the body nor within the corona, nor yet any immobile setae. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 367 Alimentary tract.— The large cavity (PL XVIII, Inf., figs. 1-3) of the corona opens externally by the ventral mouth (coronal aperture). It is without ciha and hned by a continuation of the hypodermis and its cuticula. It is succeeded by a thick-walled short oesophagus (Pi. XVIII, Oes., fig. 1), also without cilia and \nth a sphincter muscle (as Gast has described.) From the oesophagus, and attached to its posterior end, a narrow flexible tube (PL XVIII, Oes.T., figs. 1-3) extends back into the proventriculus, which may be termed the oesopha- geal tube (this was also described by Stokes for A. hucinedax). This tube is hned by a thin epithehum with a few nuclei, and its posterior shghtly enlarged end is free; it is laterally compressed, and probably possesses a musculature of its own, since it beats in rapid undulations with many changes of form ; it is very elastic to ahow the passage of the large objects of food (mainly smaller free-swimming Rotatoria, also Ostracoda, Nematoda, Infusoria and Acarina). The proventriculus (Prov.) is a large, distensible sac, of nearly the width of the trunk; its interior surface is hned by a cuticula without cilia, next fohows a clear zone of transparent protoplasm, then the peripheral cytoplasmic layer containing the nuclei. The musculature of this region described by Gast I was unable to find. The mastax Ues at the posterior end of the proventriculus; its appearance is shown in fig. 1, and the masticating teeth of one side shown in fig. 5; I have not been interested to deter- mine its finer structure, which has been done very carefully by Gast, but will simply state that its large tooth is not sharply bent at the tip as in lentiformis, and that of the four smaller teeth on each side one is fre- quently absent. Further, the usual parts may be distinguished: the unci and manubria, the fulcrum and its rami. On this follows the stomach proper (Chylusdarm, Mecznikow; oesophagus, Foulke; stom- ach, Leidy; Magendarm, Gast). This (PL XVIII, Stom., figs. 1-3) is the only portion of the tract that is ciliated, and it is the assimilative portion of the intestine, with a single epithehum of large nucleated cells containing fatty globules. The posterior intestine {P.Int.) is a wide sac lined by a flat nucleated epithehum, which opens into the dorsal side of the cloaca (CI., flgs. 2, 3), a distensible tube hned by a similar epithelium. The cloacal aperture {Cl.Ap.) is surrounded by a sphincter muscle (found by Gast) and is actually ventral, but morpho- logically dorsal since it is behind the foot. The only glands of the intestine are one pair of large stomach glands ; these {Stom.GL, figs. 1-3) are pyriform with long ducts, the body of each gland placed at the ventro-posterior margin of the posterior intes- tine, the duct curving up around the dorsal margin of the intestine to 368 PROCEEDINGS OF THE ACADEMY OF [May, join the stomach; each gland (which has been fiihy described by Gast) has one large nucleus^ or as many as three nuclei evidently produced by amitosis of the single one. The substance of these digestive glands appears granular. Foot. — This has been carefully described by Gast, and I have little to add to his description. It is immovably attached to the surface on which the animal rests by the secretion of large hypodermal glands which are apparent only in the free-swimming stage. Its outhne {F., figs. 2, 3) is more or less circular and it is placed upon the ventral sur- face of the body anterior to the cloacal aperture, its margin shghtly elevated. Musculature. — A. vorax agrees very closely with A. lentiformis in the details of the musculature, which is very strongly developed. The following muscles may be distinguished (figs. 1-3) : I. Hypodermal muscles (muscles attached at both ends to the hypodermis). (a) Circular muscles. (1) Sphincter coronse primus, Sph.C. I., a completely closed ring at the edge of the corona with a dorsal loop (Gast's rm 1). (2) Sphincter coronse secundus, Sph.C. II, a narrower muscle just behind the preceding, interrupted ventrally (Gast's rm 2). (3) Sphincter coronse tertius, Sph.C. Ill, a completely closed ring (Gast's rm 3a). (4) Sphincter coronse quartus, Sph.C. IV, present onty dorsally and not parallel to the other sphincters (Gast's mb of his fig. 2). (5) Sphincter coronse quintus, Sph.C. V, a slender muscle inter- rupted dorsally and ventrally (Gast described it as a branch of his rm 3). (6) Sphincter coronse sextus, Sph.C. VI, a completely closed ring (Gast's rm 4). (7) Sphincter coronse septimus, Sph.C. VII, the largest of the coro- nal sphincters, a closed ring giving on each side a branch to join the next (Gast's rm 5). (8) Sphincter coronse octavus, Sph.C. VIII, with a shorter dorsal and a larger ventral interruption (Gast's rm 6). (9) Sphincter trunci primus, Sph.tr. I, a small muscle, present on each side dorso -late rally only, subject to considerable variation. (10) Sphincter trunci secundus, Sph.tr. II, interrupted dorsally only (Gast's rm 7). (11) Sphincter trunci tertius, Sph.tr. Ill, interrupted ventrally and dorsally (Gast's rm S). 1903.] NATURAL SCIENCES OF PHILADELPHIA. 369 (12) Sphincter trunci qiiartus, Sph.tr. IV, interrupted dorsally but generally closed ventrally (Gast's rm 9). (13) Sphincter trunci quintus, Sph.tr., V, dividing into three branches, interrupted dorsally and ventrall}' (Gast's rm 10). (14) Sphincter trunci sextus, Sph.tr. YI, paired, latero-ventral on each side of the foot (Gast's mb 3). (6) Longitudino-circular muscles. (15) One pair, L.-C.M., passing from the posterior end of the body to about the middle of the trunk, then each turns ventrad (Gast's dim 3). (c) Longitudinal muscles. (16) Retractor coronse primus, i^.c. 7, on the antero-dorsal surface of the corona (Gast's dim 1). (17) Retractor corona secundus, R.c. II, on the postero-dorsal sur- face of the corona (Gast's dim 2). (18) Retractor coronse tertius, R.c. Ill, the innermost ventral pair, attached at the foot, passing forward and outward to insert upon Sph.C. I (Gast's vim 1, but with a different anterior relation). (19) Retractor coronse quartus, R.c. IV, paired, ventral (Gast's vim 2). (20) Retractor coronte quintus, R.c. V, inserted posteriorly at the sides of the foot, passing anteriorly to insert on Sph.C. I (it is much longer than Gast's vim 3). (21) Retractor coronse sextus, R.c. VI, paired, each with two branches ending posteriorly on the sides of the trunk, a single anterior branch inserting on Sph.C. I, and with two small lateral branches (which for the sake of clearness are not shown on the lateral view of the animal). (This corresponds more or less to Gast's vim 4.) (22) Compressor trunci, C.T., paired, subject to considerable varia- tion in form, a very thin but broad muscle always with several branches, extending back to the sides of the foot and anteriorly inserting upon R.c. Ill, R.c. IV, R.c. V. II. Visceral muscles (those which are entirely separated from the hypodermis, or in which only one end is attached to the hypodermis, which is the more usual condition). (a) Circular muscles. (23) Retractor mastacis, r.m., paired, a short muscle connecting the side of the mastax with L.-C.M. (24) Sphincter oesophagi, sph.oes. (found by Gast). (25) Sphincter ani, sph.an. (found by Gast). (b) Dorso-ventral muscles. (26, 27) Depressor trunci prinuis et secundus, d.tr. 1 and 2, two pairs, 24 370 PROCEEDINGS OF THE ACADEMY OF [May, large muscles inserted on the hypodermis a little to one side of the foot, passing dorsally through the body cavity and inserted with enlarged ends on the hypodermis on the sides of the proventriculus (Cast's dv 1 and dv 2). (28) Depressor trunci tertius, d.tr. 3, paired, smaller than the pre- ceding and placed posterior to them (Cast's dv 3). (29) Depressor trunci quartus, d.tr. 4, paired, behind the preceding (Cast's dv 4). (c) Longitudinal muscles. (30) Levator coronte, I.e., paired, inserted posteriorly on the hypo- dermis at the sides of the mastax, passing through the brain in the wall of the corona, subdividing anteriorly, and ending on Sph.C. 1, G. (Cast's Lml 1). (Cast's second visceral longitudinal muscle pair, his Llm 2, appears in vorax to be hypodermal, namely my R.C. V). (31) Defiexor coronse, dc, paired, arising dorso-laterally by two branches on the hypodermis, and passing forward to insert upon Sph.C. I (Cast's Llm 3 and Llm 4). (d) Irregular muscles. (32) Contractor corome primus, c.c. 1, lying deep in the dorsal wall of the corona, a muscle mth four arms (this is the one considered by Mecznikow to be a nerve ganglion; it corresponds to the mb of Cast's fig. 1). (33) Contractor coronte secundus, c.c. 2, a muscle ring lying deep in the dorsal wall of the corona behind the preceding, with three pairs of lateral branches, and an unpaired mecho-posterior branch (Cast's mbr). Cast has described the histology of these carefully, and I have nothing new to add to his description. Nephridial system. — This (PI. XVIII, figs. 2, 3, 6) consists of the pos- terior unpaired canal opening into the cloaca, the lateral canals, and the ductules, the latter terminating each in a flame cell. Mecznikow and Stokes have described these organs fragmentarily, and Cast very fully. The posterior canal (PI. XVIII, figs. 2, 3) which opens into the ante- rior end of the cloaca (CI.) is unpaired, thick-walled, with the lumen spirally twisted. At its anterior end the lateral canals join together, so that the terminal portion may be considered having originated by their fusion in the mid-line. Each lateral canal passes anteriorly, then bends dorsally (fig. 2), and has a very distinct lumen and a thin wall. Where the lateral canal turns dorsally it greatly enlarges in diameter to form a tripartite swelling; each of the three parts of this swelling possesses 1903 ] NATURAL SCIENCES OF PHILADELPHIA. 371 one nucleus, and each is a single large cell containing a much con- voluted, somewhat pulsatile lumen. The dorsal termination of this swollen portion lies at the boundary of corona and trunk, and at that point joins with the common nephridial ductule. From this ductule a secondary ductule passes anteriorly for a short distance, then divides into three tertiary ductules (figs. 1, 2, 6). The most median of these bears the first flame cell (Fl.C. 1), and is connected with the correspond- ing ductule of the opposite side by a transverse commissure (fig. 1) placed above the mouth cavity. The middle tertiary ductule termi- nates in the second flame cell (Fl.C. 2). The most lateral tertiary duc- tule bends ventrad, then just behind the lateral antenna divides into two quartern ductules, the anterior of which is very short and termi- nates in the third flame cell {Fl.C. 3) placed near the lateral antenna, while the posterior passes backward and follows the course of the lateral canals of the nephridia, terminating in two short ductules each ending in a flame cell (Fl.C. 4, Fl.C. 5). Finally, from the common nephridial ductule pass backward on the dorsal surface of the trunk a pair of secondary ductules, each of which terminates in a flame cell near the mastax {Fl.C. 6). Cast's account differs in that he states that what I term flame cells 4 and 5 connect by short ductules directly with the lateral canals. The nephridium of one side is thereby connected with that of the other at its anterior end by a commissure of the ductules, and at its posterior l^y fusion of the lateral canals. The only portion of the nephridia which are cihated are the terminal flame cells ; these (fig. 6) are completely closed from the body cavity, and each has an intra- cellular canal in which beats a long tuft of cilia (a typical "flame") attached at the wall of the cell where the lumen ends; the terminal end of each flame cell is somewhat enlarged, more or less amoeboid, and contains one nucleus. The lumen of the ductules and of the lateral canals is also intracellular; a pair of nuclei are always present on the wall of the commissure connecting the anterior secondary ductules. Gennarium (Eierstock, Mecznikow; Keimdotterstock, Cast; ovary of Stokes) has the same relations as in lentiforviis. It lies on the ven- tral side of the body {Ov., figs. 2, 3) anterior to the foot, often in an irregular position due to the pressure of embryos upon it, a small rounded mass of cells, with a distinct cellular lining which is continued backward as the unpaired oviduct {Ovd.) and joins the cloaca between the openings of the nephridia and the posterior intestine. It contains yolk cells (nurse cells) to the number of 10-14, large cells without dis- tinct boundaries, each with a large nucleus containing a huge nucleolus. 372 PROCEEDINGS OF THE ACADEMY OF [May, A much smaller space is occupied by the much smaller germ cells, placed near the duct, characterized by very small, deep-staining nuclei and clear cytoplasm. Gast correctly noted these two kinds of cells and their differences. The egg cells do not reach their full size within the germarium, but only when they leave it and reach the oviduct, which becomes dilated by them to serve as a uterus for the complete embryonal development ; as many as five large embryos are found in the uterus at one time, representing different stages of development. N'ervous system. — Stokes was the only one to see this organ complex before the time of the very thorough account by Gast. The cerebral ganghon lies in the mid-line above the alimentary tract (PL XVIII, Cer., figs. 1, 2), at the junction of corona and trunk or a little anterior to this point. Seen from the side (fig. 2) it appears ovoid, a little longer than high. Seen from above (Cer., figs. 1, 6) it is found to be widest in the transverse plane. Through each side of it passes a visceral longitudinal muscle (levator coronse. I.e.). In life as weh as in stained preparations the following nerves can be seen passing from this cerebral ganglion (figs. 1, 6). From its dorso-anterior edge three pairs of deli- cate nerves, which converge to the dorsal sense-organ (D.Sens.O.), i.e., pass upward and forward from the ganglion. Each of these nerves has a nucleus at the point where it joins the ganglion. From each side pass out at least eight nerves : one for the latero-anterior sense-organ (A.Sens.O); one for the latero-posterior sense-organ (L.Seyis.O.); two large nerves which pass back from the latero-posterior angle of the ganglion but which could be traced only a short distance ; and four other nerves which could be traced only a short distance, but one or two of which appeared to be connected with the nephridial tubules. The ganghon and its nerves are thus strictly bilaterally sj-mmetrical. Now the three pairs of nerves which can be traced to the three sets of sense-organs all arise from the dorsal margins of the ganghon; the dorsal portion of the ganglion may then be essentially sensory. The two large ventral posterior nerve pairs arise from two large cells, each evidently bipolar, placed at the ventro-posterior margin of the gang- lion; Gast noted only one of these and only one posterior nerve. These two cells are characterized by their great size (they might be called on this account neurochord cells, with reference to similar huge cells in Nemertini, Annelida and Crustacea), and the large size of their nuclei. Since these differ so markedly from the demonstrated sensory nerves, it is very probable that they innervate either the musculature or the viscera, i.e., that they are either motor or splanchnic in function. Then since their roots are ventral in the ganglion, it would follow that 1903.] NATURAL SCIENCES OF PHILADELPHIA. 373 there may be a differentiation within the latter of a dorsal sensory and a motor or sphlanchnic ventral portion. The nerve cells in the brain are quite numerous and symmetrically arranged, but cell boundaries are distinguishable only in the case of the two huge cells. Two con- secutive sections of one and the same ganglion are figured (figs. 7, 8)> cut in an obliquely horizontal plane, one passing through the nerves to the lateral antennse and one of the huge cells (fig. 8), and the other (fig. 7) passing further forward. In the latter can be seen what appears to be a transverse fibrous commissure on the ventral surface of the ganghon. The nerve cells are thus arranged mainly dorsally and posteriorly. There are no recognizable nerve cells on the ventral surface of the alimentary tract, as far as can be determined by the usual staining methods. Sense-organs. — The dorsal sense-organ (D.Sens.O., figs. 2, 6) is a specialization of the hypodermis on the dorso-midline of the corona? and was discovered by Gast. It consists of a group of cells (fig. 6) in which one large and five small nuclei can be seen, which compose a thickened area of the hypodermis, and above which there is a small annular thickening of the cuticula; no sense hairs were seen. The antero-lateral sense-organs, also discovered by Gast, lie each (A.Sens.O., figs. 1, 6) upon the side of the corona, as a thickening of its wall. Each consists of a cytoplasmic mass containing four large nuclei immediately imbedded in a more granular cytoplasm, while into a more hyaline cytoplasm extends a slender cylindrical rod from the cuticula; there are thus some differences from the relations in lentiformis. A single nerve, containing a nucleus at its junction with the cerebral ganghon, innervates each of these organs. The postero-lateral sense-organs (lateral antennse) lie each on the side of the corona somewhat ventral near the junction of the trunk (L.Sens.O., figs. 1-3). To each passes one large nerve, with a large nucleus at its peripheral end ; and as Gast also observed, the organ itself consists of a tubercular elevation of the body wall, and on the apex of this a cylindrical cuticular process, bearing a tuft of long, non-vibratile sense hairs (fig. 6) ; the base of the cuticular process is enlarged and extends below the hypodermis. Body cavity, connective tissiies. — The body cavity contains a clear fluid in which float masses, varying in number and size, of a brown color; these are non-cellular and probabh' represent bA'c-products of metabolism. The connective tissue elements are branched, naked cells with many processes, constantly changing their form and moving about through the body cavity. They are frequently found in close 374 PROCEEDINGS OF THE ACADEMY OF [May, connection with the walls of the nephridia, particularly with the anterior commissure. The immature female. — The free-swimming female (fig. 4), just hatched from the egg, shows the following differences from the adult. The corona is represented simply by a circular hypodermal thickening bearing a row of vibratile cilia; and in the part of the trunk projecting anteriorly beyond this ring lie a pair of large red eyes (£".). The mouth (ilf.) is ventral, and the most anterior portion of the alimentary tract contains short ciUa which do not extend out of the mouth aperture. The cerebral ganglion (Cer.) is much larger than in the adult. The whole hypodermis is thicker, and that portion of it lining the well- demarcated foot (F.) consists of about six huge cells (F.GL) which Gast has shown to be gland cells, by the secretion of which the animal firmly attaches itself. The surface of the foot disk is covered with long ciha. The musculature I have not drawn in this figure, but it is similar to that of the adult. STEPHANOCEROS, Ehrb. Historical. Stephanoceros was discovered in 1761 by Eichhorn, and he later fig- ured it (1775, according to Ehrenberg). 0. F. Miiller (1776) declared it to be a Tubularian, and Oken(1815) also placed it among the hydroids. Goldfuss (1820) first named it, calhng it Coronella fimhriata. Since this generic term was preoccupied Ehrenberg (1832) changed it to Stephanoceros, and called the species Stephanoceros Eichhornii, which all later writers have followed. According to our present generally accepted rules of nomenclature, however, and in justice to Goldfuss, this species must be named S. fimbriatus (Goldfuss) and not S. eich- hornii Ehrenb. Ehrenberg (1832, 1835, 1838) gave good figures of the animal, described the intestine, stomach glands (which all later writers seem to have overlooked), six longitudinal muscles, the ovary, and the flame cells (which he termed gills). Perty (1852) described as a new species *S. glacialis, but this is generally held to be an abnormal indi- vidual of fimbriatus. Leydig (1854) added richly to the knowledge of its anatomy and development. He was the first to note the nucleated hypodermis, circular muscles (he stated there are four longitudinal muscles in the foot which divide dichotomously), the cesophageal tube, ciha in the coronal cup and intestine, lateral canals of the nephridia (with Ehrenberg he regarded these organs as respiratory). Gosse (1855) described minutely the mastax, and in 1862 gave quite a full account of the structure; by experiment he proved the nephridia to be 1903.] NATURAL SCIENCES OF PHILADELPHIA. 375 excretory and not respiratory, found the contractile vesicle, four flame cells on each side of the body, described the dorsal glandular mass (found by Leydig) as the brain, and described five pairs of longitudinal nmscles. Cubitt (1870) found the lateral sense-organs, but made the serious error of supposing that each of them was connected with a flame cell (his pulsatile sac), and supposed them to be respiratory structures for taking water into the nephridial tubules; he carefully described the ciliary wreath within the coronal cup, but mistook (Uke Leydig and most of the later describers) the dorsal glandular body for the brain. Hudson and Gosse (1886) gave very good figures ; found five flame cells on each side of the body, described the connection of the lateral canals with the contractile vacuole, discovered the dorsal sense-organ, and described six pairs of longitudinal muscles. Vallentin studied stained sections; he found that large hypodermal cells of the foot secrete the tube ; that the dorsal mass is not nervous but glandular and opens by a duct into the vestibule; he regarded the large hypodermal cells at the bases of the arms to be nerve cells. The male of this species has been described by Western (1893) and more carefully by Dixon- Nuttall (1896). Less important are the papers of Dujardin (1841), Weisse (1845), Dobie (1854), Pritchard (1861), Cubitt (1869), Peirce (1875), Rosseter (1881, 1884), Lord (1885), Jennings (1894, 1896, 1900, allfaunistic). Anato:mical. Nothing new can be added to our knowledge of the external jorm (PL XIX, fig. 16). The margin of the corona is prolonged into five long arms, which are slightly fiattened and extensile. These arms are shghtly curved, and the dorsal, unpaired one slightly longer than the others. The body is long and slender, rather strongly demarcated from the long foot which ends in a short peduncle (fig. 14). The gela- tinous tube (Tub.) is very transparent, with annular folds, elastic and of great thickness. Hypodermis. — The hypodermis {Hyp., figs. 9-11) of the trunk is a thin layer with flattened nuclei. At the margin of the corona it is much thickened, so as to form at the base of each of the arms a mass of four or five large rounded nucleated cells. The hypodermis of the arms is about as thick as that of the trunk, but neither in the living state nor upon preparations in which the nuclei of other regions of the hypoder- mis are stained very sharply are nuclei to be found in it. The hypo- dermis of the arms of the corona would therefore appear to be a direct continuation of the cytoplasm of the large hypodermal cells at the 376 PROCEEDINGS OF THE ACADEMY OF [May, base of the arms. Each of these cells then has a great extent, from its base to the free tip of an arm. When the arms are fully extended the outlines of the hypodermis seen on optical section appear even, without any irregularities. But when the arms are partially retracted, the hypodermis appears partitioned into circular areas on its inner surface; on surface view these are seen as in fig. 13; and on optical section as in fig. 12. At first I was inclined to consider these as representing component small cells of the hypodermis, but the absence of nuclei in them excludes that view and they can be regarded only as folds of the continuous hypodermis due to the contraction of the arm. The arrangement of the cilia upon the arms is very complicated and has not yet been satisfactorily determined; it is best seen on a living arm in a contracted condition. On the outer as well as upon the inner surface of each arm the arrangement of the cilia is shown by the dotted lines in fig. 12; there the cilia are arranged in single rows, making angles together, these single lines of ciha separated by spaces which are without cilia. But on lateral view each arm shows a different arrangement of the cilia; here (fig. 13) the cilia are arranged in tufts placed on oblique thickenings of the surface of the hypodermis. In this figure only those cilia seen in profile are drawn in extenso, while the insertion points of those on the aspect of the arm toward the observer are represented by stippling, each patch of fine dots denoting a tuft of cilia. At the tips of the arms the ciha are more evenly arranged along the whole hypodermis. The appearance of the tufts of cilia, arranged in oblique bands, are shown for the two ventral arms under a lower scale of magnification in fig. 16. The cilia are extensile, and in fully expanded condition attain a length considerably greater than that of the arms themselves; but this is to be seen generally only when the animal is under the pressure of a cover- glass. I could not determine the presence of a cuticula upon the arms. The hypodermis of the foot is thickened, and at the junction of this region with the trunk (figs. 9, 10) certain cells are very prominent. These may correspond to the foot glands of the other Rotatoria, and secrete the substance of the tube in which the animal lives. The short peduncle (fig. 14) by which the foot is attached probably represents a secretion of hypodei'mal cells at the distal end ; this peduncle is short, hard and homogeneous. Alimentary tract. — By an oblique diaphragm {Dia., figs. 9, 10) the cavity of the corona is divided into a more spacious infundibulum (Inf.) and a more posterior vestibulum (Vest.), both of which have a nucleated lining similar to that composing the hypodermis. The diaphragm is 1903.] NATURAL SCIENCES OF PHILADELPHIA. 377 an annular fold composed of a more thickened cell layer; it is shown in oblique ventral view in fig. 11, on lateral view in fig. 9, and on dorsal view in fig. 10. The drawings show that this ring fold bears ciha along the anterior border of its ventral half, a tuft of longer ciha on each side, while the margin of its dorsal half has no cilia. Only this diaphragm bears ciha, and none are found upon the walls of the infundil^ulum and vestibulum. By a narrow aperture there follows upon the vestibulum the oesopha- geal tube {Oes. T.), the distal end of which projects freely without attachment into the cavity of the proventriculus. This is an elastic tube that beats rapidly in many dilations and therefore probably pos- sesses a musculature of its own; I could not determine nuclei in its wall. The proventriculus (Prov.) is the largest portion of the ahmen- tary tract, its wall composed of a single epithelium of large cells with distinct nuclei, and lined internally by a cuticula. In this posterior part lies the mastax, concerning which it will suffice to refer to fig. IS, a dorso-posterior view. Gosse was wrong in considering it to be rudi- mentary, for it has well developed unci and manubria, fulcrum and rami. The stomach (Stom.) is a rounded, dorsally placed sac lined by large cells which bear numerous long ciha. The posterior intestine (P.Int.) follows, lined by a flattened epithelium without cilia and with distinct nuclei; and behind it a very short rectum opening into the cloaca (CL). A single pair of large stomach glands {Stom.GL, figs. 9, 10) are present, each latere- ventral at the junction of proventriculus and foot, and each containing several nuclei. Musculature. — The following longitudinal visceral muscles run from the foot through the body cavity of the trunk to attach themselves to the bases of the arms of the corona (figs. 9-11). One pair to the dorsal arm, extending to its outer sphincter muscle. A pair of dorso-lateral muscles, each of which inserts anteriorly on the thickened hypodermal ring of the corona at the junction of the dorsal with a lateral arm. And a pair of ventro-lateral muscles, each of which, when reaching the thickened hypodermal ring of the corona, divides into two branches which pass forward to the outer sphincter muscle of the coronal arms. Thus I find only three pairs of longitudinal muscles. All these muscles have their muscle bodies with nuclei contained in the foot. Of the circular hypodermal musculature, there are a large number of dehcate sphincters in the foot and trunk region, and at the anterior end of the corona two sphincter muscles of large size (Sph.C. I and //, figs. 9-11). The remaining complicated network of musculature of the corona I have not studied. 378 PROCEEDINGS OF THE ACADEMY OF [May, Nephridial system. — The whole iiephridial system of one side of the body is shown in fig. 9, and a portion of it from dorsal view in fig. 10. Into the posterior end of the cloaca (CL)opens the unpaired contractile vacuole {Con.V.); and into this a long elongate unpaired sac with dis- tinct outer lining and containing a convoluted tube with wide, pulsatile kmien, the common terminal canal. At the free end of the latter unite with it the two longitudinal canals of the nephridia. Each of these passes forward latero-ventrally to al^out the level of the stomach glands, and there becomes much convoluted so that the exact course is difficult to follow; but through all these convolutions it remains a single tube with distinct walls. Each lateral canal then turns dorsad to about the position of the cerebral ganglion where it again Ijecomes convoluted ; its diameter is greatest in this portion of its course. At this anterior end the ductules are given off to the flame cells, and, as fig. 9 shows, one ductule passes forward and divides into five ductules, each ending in a flame cell {Fl.C. 1, Fl.C. 2). A second ductule passes posteriorly and ends in a flame cell (Fl.C. 6) placed near the stomach. In one individual I saw two other flame cells {Fl.C. 7, Fl.C. 8) placed near the sixth flame cell, but their connections with the ductules of the nephridial system I did not find. As the dorsal view, fig. 10, shows, in front of the cerebral ganglion a transverse commissure connects the ductules of the two sides of the body. Each flame cell (fig. 15) bears internally a flame of long cilia, and appears completely closed from the body cavity. The lumen of all the ductules and the main canals is intracellular, and cilia are found only in the flame cells. The whole nephridial system is exceedingly distinct in life, provided the animal is well expanded. Dorsal glandular organ. — On the dorsal side, at the junction of the corona with the trunk, lies a large mass (D.Gl., figs. 9, 10) of clear vacuolar structures, which has been heretofore regarded by most writers as the brain. It is composed of clear transparent globules, varies in size in different individuals, and becomes shrunken by the action of alcohol. Nuclei are to be seen at its margins, particularly posteriorly, but I was unable to determine whether each of its com- ponent globules represents a cell. This organ is closely attached to the dorsal hypodermis around the dorsal sense-organ {D.Sens.O., fig. 9), and is penetrated by the nerves which pass to the latter from the cerebral ganghon (Cer.) ; it has no connection with the alimentary tract. The functional significance of this organ is whoHy obscure, and I have supposed it to be glandular simply on account of its con- nection with the hypodermis and the dorsal sense-organ. It is not 1903.] NATURAL SCIENCES OF PHILADELPHIA. 379 probable that it forms the secretion for the tube, since it is of very large size in those large individuals with fully-formed tubes. Germarium, oviduct. — The germarium {Ov., figs. 9, 10) is a more or less spherical organ placed on the ventral side of the body, composed of a syncytium (fig. 17) containing fully fifteen large nuclei which are nurse cells (yolk cells), and with at one end much smaller nuclei (ovogonia). The whole is surrounded by a nucleated membrane which is continued as a thin-walled unpaired oviduct (Ovd.) to open into the ventral sur- face of the cloaca between the rectal and the nephridial openings. One ovum matures at a time within the germarium, then is discharged into the oviduct which serves as a uterus for its further development until it reaches the free-swimming stage. As many as four or five embryos are to be found at once in the uterus. Nervous system. — Most of the preceding writers have mistaken the dorsal glandular body for the ganglion, while Vallentin recognized this body as glandular, but supposed it to open by a duct into the alimentary canal, and considered the large hypodermal cells at the bases of the arms to be nerve cells. The cerebral ganglion (Cer-., figs. 9, 10) lies directly below the dorsal glandular mass, is elongate from side to side and approximately cylindrical on lateral view. On stained prepara- tions I could determine that it is composed of many small nerve cells, the nuclei of which stain deeply. On lateral view of the expanded living animal (fig. 9) the following nerves are seen to arise from it : (1) One pair which passes directly dorsad or dorso-caudad to innervate the dorsal sense-organ (D.Sens.O.) ; this pair penetrates the dorsal glandular body, and probably represents what Vallentin mistook for a glandular duct leading into the vestibulum. (2) A pair of nerves to each lateral sense-organ (L.Sens.O.). (3) An unpaired, very delicate median nerve to the dorsal hypodermis behind the dorsal sense-organ. (4) A pair of nerves which pass backward upon the sides of the prov- cntriculus (Prov.). And (5) a pair of nerves, the largest of all, which pass ventrad on the sides of the proventriculus. I have found no evidence that the large hypodermal cells at the bases of the arms are nervous; the long fibres which Vallentin found proceeding from them probably represent the continuations of these cells to form the walls of the arms. Sense-organs. — Eyes were seen only in one individual, a pair of small red spots (E., fig. 9), lying close to the cerebral ganglion. The dorsal sense-organ {D.Sens.O., figs. 9, 10) is a slight thickening of the hypo- dermis bearing a tuft of short sense hairs, which penetrate a circular aperture of the cuticula. Each lateral sense-organ (L.Sens.O.) has 380 PROCEEDINGS OF THE ACADEMY OF [May, one conspicuous large nucleus, and likewise bears a tuft of short cilia. The body cavity is voluminous in the trunk and the foot (but does not continue into the peduncle of the latter). In it are free-floating, non-cellular corpuscles, probably waste products. The cavity of the arms is continuous with that of the trunk (figs. 9-11). Very few connective tissue cells are present in the body cavity. FLOSCULARIA, Oken. Historical. Of the numerous writers upon this genus only the follo^\'ing have considered to any extent the internal anatomy : Ehrenberg (1834. 1838) Avi'ote with all the enthusiasm of the naturalist: "Sie sind einzeln, wie seltene schone Blumen auf einer Wiese, und erfreuen gleich ihnen das Auge des Beobachters." He found the mastax, oesophageal tube and the stomach glands. Dobie (1849) gave a careful description with good figures, and distinguished the infundibulum, yestibulum and proventriculus ; he saw the cilia on the diaphragm. Leydig (1854) added to this description in finding the contractile vacuole and the longitudinal muscles. Gosse (1855) described in some detail the masti- catory apparatus; and (1862) erroneously termed a vascular system the narrow spaces of the body cavity in which float excretory cor- puscles. The account of Moxon (1864) is much more thorough than any of the preceding, and contains the discovery of the dorsal and lateral antennae, correct position of the cilia within the infundibulum, and occurrence of the nephridia with four flame cells on each side of the body. Cubitt (1869) studied particularly the cilia within the coronal cup. The papers of Grenacher (1869) and Cubitt (1872) added little new. Eckstein (1884) gave a careful account of the hypodermis, and the muscles and peduncle of the foot. Hudson and Gosse (1886) add but little to the account of Moxon. The paper by Hood (1895) I have not seen. The males have been described by Hudson and Gosse and by Weber (1888). But the best of these descriptions, those of Dobie and Moxon, leave much untouched in regard to the finer anatomy. Anatomical. F. campanulata Dobie (PI. XX, figs. 27-35; PI. XXI) was the species obtained in the greatest numbers, and on this account its anatomy could be more fully determined than that of the others. F. conklini Montg. (figs. 25, 26, PI. XX) was obtained at the same time, but it is rather a difficult form to investigate on account of the large number 1903.] NATURAL SCIENCES OF PHILADELPHIA. 381 of minute floating corpuscles within the body cavity which greatly ob- scure the internal organs. F. coronetta Cubitt (PL XX, figs. 22-24) and F. ambigua Hudson (PI. XIX, figs. 19-21) were the other species studied, but ambigua in only a few individuals, so that for this species my observations are unfortunately fragmentary. General form. — In all Floscularians there may be distinguished the corona, trunk and foot (PI. XXI, fig. 37). The corona is in ante- rior extension and enlargement of the trunk. In campanulata (PL XX, figs. 27, 28, 31) it is largest, widely bell-shaped and prolonged into five broad lobes, the dorsal of which is largest and longest, next the two ventral lobes, and smaller than these the two lateral lobes. In this species the corona is subject to considerable individual variations in form, but is always much wider and fully as long as the trunk. In ambigua (PL XIX, fig. 20) the corona is relatively smaller, and of its five lobes the lateral pair are usually very small and often distinguish- able only by the cilia which they bear. In conklini (PL XX, figs. 25, 26) the corona is more cylindrical, frequently considerably less than half the length of the trunk, the lateral lobes also very small. In coronetta (PL XX, figs. 22, 23) the lobes are more cylindrical and narrowed, with enlarged (knobbed) tips, the dorsal one only slightly longer than the others. The trunk is elongate and cylindrical, in conklini (fig. 25) more arched on the ventral than on the dorsal side. The foot is an elongate proxi- mal extension of the body, relatively longest in conklini and coronetta. The foot ends in a peduncle by which the animal is firmly attached to the surface of a water plant stem. This peduncle is much longer than broad in campanulata (PL XXI, fig. 40) and cmibigua, little longer than broad in coronetta (PL XX, fig. 24), and fully as broad as long in conklini (fig. 26). These different species were all found attached to MyriophyUuni, but to different parts of it, campanulata always to the very finest outgrowths of the plant; conklini usually to the angles at the bases of stems; ambigua and coronetta to larger stems. In all these forms the animal is surrounded by a gelatinous, elastic tube (PL XXI, fig. 37, Tub.), with an anterior opening; in conklini (PL XX, fig. 26) the tube is relatively largest and its outer surface often covered with foreign particles, and in campanulata it is exceed- ingly transparent and visually without any such adherent particles. By a weak solution of methylene blue the tube quickly becomes colored intensely; and this method of demonstrating it shows its form with great distinctness. The animals are highly contractile, thanks to the strongly developed 382 PROCEEDINGS OF THE ACADEMY OF [May, longitudinal muscles; the coronal lobes may be folded into the infun- dibulum, the whole corona and trunk much shortened, and the foot so contracted that its peduncle is brought close to the trunk. Hypodermis. — The hypodermis of the trunk is a thin, one-layered epithehum, ^vith relatively few nuclei (Hyp., figs. 20, 22, 23, 25, 27, 28). It bears a hyaline, structureless cuticula (Cut.). On the free margin of the corona it is much thickened and contains large nuclei which are easily seen even in life. In campanulata (PI. XX, figs. 27, 28) this coronal thickening follows the margin of each lobe, and at the tips of the dorsal and ventral lobes are particularly large nuclei. That is also the case with amhigua (PL XIX, fig. 20). In conklini (PI. XX, fig. 25) the dorsal lobe has the hypodermis thickened only at its tip. In coronetta (figs. 22, 23) the hypodermal thickening forms a ring around the base of the lobes, while the hypodermis forming the walls of the lobes is but little thicker than that of the trunk; this is an approach to the condition in Stephanoceros , and wall probably be found to be the case in all forms where the lobes are slender. Another peculiarity of coronetta is that the cuticula covering the lobes is thicker than else- where upon the body ; in the other species I could not distinguish with certainty a cuticula upon the lobes, though probably a deHcate one is present. The cilia of the external surface are limited in the adult to the mar- gins and lobes of the corona. In conklini (figs. 25, 26) these cilia are no longer than the corona, and strongly vibratile ; they are found also only upon the tips of the lobes. In coronetta the tip of each coronal lobe bears a tuft of long cilia (fully as long as the trunk), which have a slow undulatory movement; while the sides of the lobes and the interlobular margin of the corona bears much shorter cilia with a more rapid movement. In campanulata (PI. XXI, fig. 37) the ciha fringe the whole free margin of the corona. Here they are longest upon the tips of the lobes, but all have a length greater than that of the corona and trunk together, and sometimes nearly two-thirds that of the entire animal. The cilia in this species are non- vibratile, or at most with slight undulatory movements, and toward their free ends become grad- ually exceedingly tenuous. All preceding authors have figured them much too short, probably because their full length can be determined only with high powers of magnification. These cilia are less hke true cilia than like the tenuous, stiff pseudopodia of Heliozoa; and hke the latter they are extensile and retractile to some degree at least, as is shown by a protoplasmic flowing. They serve not to obtain food par- ticles by ciliary currents, but rather as a wide basket to encompass the 1903.] NATURAL SCIENCES OF PHILADELPHIA. 383 prey. Generally they are held stiff and rigid; but at times wave-hke undulations are seen to pass slowly along them. Unhke a heUozoan pseudopodiuni there is no supporting axial filament within them. Ambigua has cilia like those of campanulata, and also fringing the whole coronal margin. In these two species and in coronetta the ciha project out of the anterior end of the bodj^ in a thick tuft when the animal is strongly contracted. In the foot the hypodermis is thicker than in the trunk, composed of larger cells. Its cells are largest at the junction of the trunk with the foot (PI. XIX, fig. 20; PI. XX, 22, 23, 25, 27, 28); and these may be gland cells which produce the gelatinous tube. The peduncle (figs. 24, 40) is homogeneous without cells or nuclei ; it probably represents a secretion of certain large hypodermal cells found in the embryo but absent in the adult, which are placed at the distal end of the foot (PL XXI, fig. 38). The peduncle is firmly attached to the plant stem, so securely that when the animal is torn loose from its tube the peduncle is often left still attached. Alimentary tract. — The following regions are marked in all the forms : infundibulum, vestibulum, oesophageal tube, proventriculus, stomach, posterior intestine, rectum, cloaca. The infundibulum {Inf., figs. 20, 22, 25, 28) is the cavity of the ante- rior portion of the corona, and its opening to the exterior constitutes the broad mouth. It is lined by a thin one-laj^ered epithelium in which nuclei can be distinguished by staining, is without cilia, and its wall is separated from the hypodermis to which it is attached only at its anterior margins, by the body cavity. The infundil)uhun is partially bounded off from the vestibulum behind it by the diaphragm. This {Dia., figs. 22, 23, 25, 27, 28) is a thickened annular fold of the internal wall of the alimentary tract, and the only portion of the latter anterior to the stomach which is ciliated. Viewed from in front, i.e., from the mouth opening, it appears circular with a knob-like thickening at each side in campanulata (fig. 31). Each of these knobs bears a tuft of long and slowly vibratile cilia, while the border of the diaphragm ventral to these knobs bears a semicircle of very short and delicate, rapidly vibratile cilia. There are no cilia dorsal to the knobs. In conklini (fig. 25) the relations are in general similar. In coronetta (figs. 22, 23) the dorsal free margin of the diaphragm is considerably thicker, and bears short pointed projections (like those within the infundibulum of Apsilus) ; while the ventral margin bears rather long cilia upon sev- eral knobs. In all these species (I did not determine the relations in ambigua) the cilia appear usually to beat forward, i.e., into the in- 384 PROCEEDINGS OF THE ACADEMY OF [May, fundibulum, but frequently they point backward and beat in the vestibuhun. The latter (Vest., figs. 20, 22, 23, 25, 27, 28) is the portion of the alimentary tract placed at the base of the corona, at the junction with the trunk; its lining is like that of the infundibulum. Attached to the narrowed posterior aperture of the vestibulum is a slender tube, the oesophageal tube, (Oes.T., figs. 20, 22, 23, 25, 27, 28), which extends back without posterior attachment into the proven- triculus. This tube is very thin-walled, and constantly moves in rapid undulations and distortions, so that its shape can be made out only when it comes to rest. Nuclei could not be determined in its walls in life, and all fixatives preserve it as poorly as they do the nephridia; so that stained preparations did not help in understanding its structure. But very delicate spirally arranged muscle fibrils compose a part of its wall; and it is probable that it possesses a nerve center of its own since it continues to beat after the other organs have ceased to live. In a state of rest (figs. 34, 35) it is cylindrical with a posterior enlarge- ment, and a very small posterior aperture into the proventriculus. It is very elastic to allow the passage of large objects of food {Infusoria and Mastigophora in the species studied), and may sometimes be everted anteriorly into the vestibulum. The proventriculus (Prov., figs. 20, 22, 23, 25, 27, 28) is a large sac lined by a rather thick epithelium, with readily distinguishable nuclei; its internal surface is bounded by a delicate cuticula. At its posterior end lies the mastax (Mast.). This consists of chitinous, jointed parts, the manducatory apparatus, and of musculature for their movement. Gosse (1855) holds that Stephanoceros and Floscularia have no mastax in the sense that this term is used for other Rotatoria, i.e., that in the Flosculariichv. it is not a special portion of the alimentary tract with a lining of its own. But it is surely homologous with that of the other groups, and though relatively very small I find it is not rudiment- ary but possesses all the characteristic parts found in the other Rotatoria. Thus in campanulata (PI. XXI, fig. 39) the manubrium consists of two teeth, together constituting the uncus (Un.), and of a basal piece, the manubrium (Man.). A broad chitinous plate belonging also to the malleus lies below the uncus. The incus consists of a median fulcrum (Fid.) and of lateral rami (Ram.). The muscula- ture is difficult to determine in its exact arrangement, so that I have not drawn it, but is readily seen in the living animal. The organ then has a special lining, the musculature, and is truly a mastax in the sense of Gosse, even though it does not form a separate compartment of the alimentary tract; and this is the case also in Stejjhanoceros and Apsilus. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 385 The stomach (Stom., figs. 20, 22, 25, 27, 28) is thiclc-waUed, with large cubical cells bearing long ciHa ; this is the assimilative portion of the intestine, and its cells are more or less filled with globules. Upon it follows the posterior intestine (P. Int.), hned by a thin-walled, nucle- ated epithelium without cilia. Here the faeces form large food balls in ambigua, but not in the other species. By a short rectum (Rec), often barely distinguishable from the posterior intestine, an opening into the cloaca (CI.) is attained. The cloaca is ciliated apparently only in coronetta (figs. 22, 23), is very thin-walled, and opens to the exterior by the cloacal aperture placed dorsally at the junction of the trunk and the foot, or a httle anterior to it. The diameter of the different por- tions of the alimentary tract is dependent upon the amount of food contained in them. One pair of stomach glands (Stom.GL, figs. 20, 22, 23, 25, 28) is pres- ent in all the species, placed latero-ventrally at the boundary of stomach and proventriculus. In coronetta (figs. 22, 23) and ambigua (fig. 20) each gland is elongate with a single large nucleus, in conklini (fig. 25) pyriform with several nuclei, in campanulata (fig. 28) of the same shape but wuth one nucleus. The connection of these glands with the stomach was determined positively only for ambigua, but there can be doubt that they discharge into the stomach and not into the proventriculus in the other forms also. A peculiar structure was found in every individual of ambigua, but in none of the other species. This was a body of a brown color (X.Y., fig. 20) placed in the posterior intestine, with a thick wall, a deep-staining body (nucleus ?) in this wall, and rather vacuolar contents. It appeared to lie within the posterior intestine. It can hardly be a parasite, since exactly one of these bodies was found in the same region in each individual. Musculature. — An exhaustive study of the musculature, such as was given for Apsilus, was not attempted for Floscularia. Of the hypo- dermal musculature (the following description applies particularly to campanulata) a number of delicate transverse muscles are present around the foot and trunk (fig. 27) ; by their contraction these portions of the body are compressed, and the body fluid driven forward to unfold the corona. In the corona this musculature is richly developed (figs. 27, 28), though the separate muscles are fine. Two sphincter muscles are present here, one close to the coronal margin and another behind it. Numerous slender longitudinal muscles connect these to- gether and with the hyopdermal thickening. The diaphragm has its own sphincter, and receives branches of the lateral and ventral longi- 25 386 PROCEEDINGS OF THE ACADEMY OF [^lay, tudinal trunk muscles; while the walls of the vestibulum have a rich network of muscles. There are three pairs of large visceral longitudinal muscles (figs. 22, 23, 25, 27, 28), a pair of dorsal {D.M.), of lateral (L.M.) and of ventral (V.M.) muscles respectively, attached anteriorly to the coronal margin, passing the whole of the trunk and in the foot converging together to form a muscle band which extends posteriorly as far as the proximal end of the peduncle (fig. 40); the cell bodies (Musc.C.^ fig. 28), and nuclei of these large muscle cells are placed within^the foot. Probably the proventriculus has its proper muscula- ture, thought it could not be determined ; and the muscles of the oesoph- ageal tube have been already mentioned. Nephridial sijstem.— This is essentially the same in conklini (fig. 25), campanulata (28, 30, 41) and coronetta (figs. 22, 23). Opening into the posterior end of the cloaca is an unpaired, thin-walled contractile vesi- cle (Con.V.), and into this an unpaired contorted canal enclosed by a membrane. From the proximal end of this canal pass forward the two longitudinal canals, one on each side of the trunk, a httle more ventral than dorsal in position. The anterior portion of each of these canals becomes convoluted, and tliis convoluted portion is thicker than the rest of the duct with a narrower lumen, and somewhat contractile. It is only by continued study of compressed hving individuals that the bendings and interlacings of these very transparent canals can be deter- mined. Anteriorly each longitudinal canal gives off an anterior and a posterior ductule. Each anterior ductule divides close to the lateral wall of the corona into three secondary ductules, each terminating in a single flame cell (Fl.C). The posterior ductule passes caudad to about the region of the oesophageal tube on the side of the body, then divides into two secondary ductules, a shorter dorsal one ending in a flame cell on the side of the mastax, and a longer ventral branch ter- minating in another flame ceU on the latero-ventral aspect of the trunk. In conklini I could not ascertain the mode of union of the ductules with the lateral canals, on account of the large number of those floating corpuscles within the body cavity which greatly obscure the nephridial organs. F. ambigua (fig. 21) differs from the preceding species in having six flame cells and ductules on each side of the body, i.e. an additional one near the anterior end of the longitudinal duct. Amhigua also differs from the other forms in having the membranous tube which encloses the posterior unpaired canal much longer than in the other species; but I did not have sufficient material to determine its exact length. Only in coronetta (fig. 23) is there an anterior com- missure joining branches of the anterior ductules of the two sides of 1903.] NATURAL SCIENCES OF PHILADELPHIA. 387 the body. In Floscidaria the flame cells are readily found, but it is very difficult to find all the connections of the ductules and tubules. Cilia are found only within the flame cells, as a long intracellular tuft or flame, and these terminal cells appear entirely closed off from the body cavity; they have also no connection whatsoever with the lateral antennae (as one observer had supposed). Organs of prohlcmatical significance.— In conklini there is found a large transparent structure {GL, fig. 25) immediately below the hypo- dermis (and peripheral to the other internal organs), at the junction of the corona and trunk. It is lobed and forms here a ring just beneath the hypodermis, and attached to it. It appears to be gelatinous. Prob- ably this is a glandular structure, and the dorsal gland of Stephano- ceros may be homologous. In coronetta a vesicular dorsal, hypodermal structure (D.GL, fig. 23). probably a gland, is found just below the dorsal sense-organ. Nervous system, sense-organs.— In all the species the cerebral gano-li„n (Cer., figs. 20, 22, 23, 25, 28) lies deep below the hypodermis, on the dorsal surface of the vestibulum; it is elongate from side to side and composed of numerous small nerve cells. The nerves arising from it were studied most thoroughly in campanulata. In this species"(figs. 28 32) the gangUon seen from the side shows quite clearly a distinction between a dorsal and a ventral portion. From the dorsal side two nerves pass anteriorly to the dorsal sense-organ (D.Sens.O.), and a single nerve in the median line posteriorly to end on the hypodermis. From the anterior edge of the ventral portion of the ganglion a laroe nerve fiber passes antero-laterally to each lateral sense-organ (L.Sci^s. 0.) ; two smaller nerve pairs pass posteriorly from its posterior margin'; and from its lateral margin a pair of large nerves on each side down- ward and backward along the surface of the proventriculus. These nerve fibers could be seen more clearly in fife than upon stained prep- arations; methylene blue staining gave no positive results. In coronetta there is found on the sides of the corona a single pair of small red eyes {E. , figs. 22, 23) . In some adult individuals o'f ambigua (fig. 20), but not in all, were found a number of eyes, up to five in all on the dorso-posterior part of the corona; these have an irregular ar- rangement, and each consists of a few large red pigment granules. Campanulata and conklini have no eyes in the adult stage. ^ All the species possess an unpaired dorsal {D.Sens.O., figs. 20, 22, 25, 28) and a pair of lateral sense-organs {L.Sens.O., figs. 20, 23, 27, 28) (antennse) upon the corona, each bearing a tuft of long sense hairs. In campanu- lata the dorsal organ {D.Sens.O., fig. 32) was found to consist of from 388 PROCEEDINGS OF THE ACADEMY OF [May, four to six hypodermal cells, the sense hairs of which project outward through an aperture in the cuticula. Female genital organs. — In campamilata (Ov., figs. 27, 28), coronetta {Ov., figs. 22, 2.3) and conklini (fig. 25) there is a more or less rounded germarium upon the ventral surface of the trunk; its cellular lining is continued caudad as the oviduct (which serves also as an uterus), and the latter opens as an unpaired tube into the cloaca between the openings of the rectum and the nephridia (figs. 29, 30) . The germarium consists for the greater part of its bulk of a syncytium of yolk cells with large nuclei (Yk.N.) and huge nucleoli; there are some fifteen or more of these cells. At one end of the germarium is a cap of cells with much smaller nuclei, which are ovogonia (Ovg.). As the latter increase in size they are pushed in succession into the oviduct, where the cleavage commences. In conklini and campanulata two or three large ova are found in the oviduct (uterus) at once, in caynipanidaia never more than a single one. In ambigua the germarium has an entirely different form (Ov., figs. 19, 20); it begins proximally upon the dorsal left-hand side of the trunk, extends down that side, then across the ventral region to the right hand of the trunk, and at the latter point the mature ova are found in the oviduct. I could not determine, owing to lack of material, where the ovogonia are placed within this remarkable germarium. Body cavity. — This lies beneath the hypodermis, is continued in the coronal lobes, and in the foot as far posterior as the peduncle. In it float masses of minute, brownish, non-cellular corpuscles, which vary in number in different individuals of the same species, and in conklini are always exceedingly numerous and very minute. They are dissolved by alcohol, and the larger of them often appear doubly refractive. They must be metabolic products, probably waste products, but I have never found them within the nephridial lumina. In ambigua when the animal is fully extended and somewhat compressed in life beneath the cover-glass, it appears as if these corpuscles flowed within special channels between the hypodermis and the infundibulum, and like Gosse (1862) I was at first inclined to beheve that there existed here a subhypodermal vascular system. But further study showed that the apparent canals are not fixed structures but simply portions of the general body cavity. The immature female. — This stage was studied particularly in ambigua, but the general characteristics are the same in all the species. This free-swimming stage (PL XIX, fig. 21), just hatched from the egg, has only an incipient corona and foot. The other differences from the 1903.] NATURAL SCIENCES OF PHILADELPHIA. 389 adult arc mainly the following: Two red eyes (E.) are present in all the species except in conklini. No coronal lobes are yet developed, but at the anterior end of the trunk is a thickened hypodermal ring bearing a row of vibratile ciUa. On the ventral side of and just behind this ring is the mouth opening (M.), leading into a capacious infundibu- lum (from which a vestibulum has not yet become demarcated), the inner lining of which bears long cilia. It is important to note that these cilia lie within the alimentary tract, i.e., belong to its inner lining and thus cannot represent a cingulum. The thickened hypodermal ring becomes subsequently prolonged to form the coronal lobes, and the cilia of the infundibulum are at no time a portion of the corona. Above this portion of the alimentary tract lies a large dense mass, part of which may represent the cerebral ganglion (Cer.). The hypodermis of the foot consists of a few enormous cells. A tuft of cilia projects posteriorly from the posterior end of this incipient foot; and in its axis lies a hollow cylinder of cells enclosing a granular, elongate body, which may represent the gland (F.Gl.) which forms the peduncle. The tuft of cilia at the posterior end of this embryonic foot was found, in a somewhat older stage of campamdata which had attached itself and developed a peduncle, to be still present and placed at the junction of the foot and peduncle; in the figure (PL XXI, 38) the cylindrical cellular mass within the foot probably represents embryonic muscle cells of the foot. The remaining organs are essentially as in the adult. The 7nature male of F. companulata. — The males of this species were found in November and the first half of December; the male eggs, as in other Rotatoria, are more numerous and smaller than the eggs which give rise to females. The mature male (PL XXI, fig. 36) in size and general structure, disregarding the sexual organs and the ali- mentary tract, shows a great similarity to the immature female. It lacks an alimentary tract entirely, and I was unable to find a nephridial system, though undoubtedly the latter must be present. The thick- ened hypodermal ring at the anterior end of the body bears a single ring of long, vibratile cilia, and in the projection of the trunk anterior to this ring lie two semicircular, dorsal, red eyes (E.). The foot (F.) is very short, without peduncle. The hypodermis (Hyp.) is thickened and with an irrcgurlaly scalloped inner contour. A dorsal sense-organ (D.Sens.O.) is well developed. In the anterior region of the trunk lies a large mass which may in part represent a cerebral ganglion (Cer.). The genital organs consist of a huge sperm sac (Sp.S.) connected with a cirrhus (Cir.). This sperm sac is filled with spermatozoa, and its walls thin except at one point on its dorsal surface which is thickened ; 390 PROCEEDINGS OF THE ACADEMY OF [May, probably this thickening represents the germinal epithelium and there- fore the testis {Test.) proper. A curled thick-walled tube, the cirrhus, follows the sperm sac; its lumen, the vas deferens (V.D.), is very narrow. A thin-walled short tvibe connects the posterior end of this cirrhus with the dorsal genital aperture; and wdthin this tube beat long cilia which are attached to the posterior end of the cirrhus. The cirrhus may be protruded some distance out of the genital aperture, and probably serves as an intromittent copulatorj^ organ. On the ventral side of the cirrhus, in close attachment to its wall, is a large dense body (Gl.) with an axial pyriform clear space; I interpret this to be a gland, and the clear space to be its duct. Just posterior to this gland is a lobed body projecting into the body cavity, bearing on one of its surfaces long cilia which beat in the body cavity. No further structures were observed in the living animal. The copulation was not observed. In a number of the females studied at the time the males occurred were found spermatozoa, from one to about a dozen in each female. All seen were within the body cavity of the female and none in any portion of her genital tract. The spermatozoa (PI. XX, fig. 33) are relatively huge lumbricoid cells, very elongate, the more vibratile end of which appears to correspond to a thick flagellum; they are exceedingly active and twist about within the female, and within the sperm sac of the male, like animated cork- screws. CoxcLUDixG Remarks. The family of the Flosculariidce includes the three genera Floscularia, Stephanoceros and Apsilus. Leidy's description of Acydus is too incom- plete to allow us to determine whether this form should be included in the group also. Apsilus differs quite markedly from both Floscu- laria and Stephanoceros (these two are very similar), but agrees with. them in possessing an cesphageal tube, in the similar relations of the nephridial and nervous systems, and in the great similarity of the young. The immature females just hatched from the egg have the foot bearing a tuft of ciha at its distal end. The general characteristics of the family are as follows, based upon the few species already known anatomically : The anterior portion of the trunk is expanded to form a capacious coronal cup, the free margin of which is prolonged into lobes {Stephanoceros and most species of Floscularia), or is without lobes {Apsilus, Floscularia edentata Collins and pelagica Rousselet). The whole free margin of the corona con- stitutes the boundary of a large mouth aperture. Cilia are absent upon the corona in Apsilus. In Stephanoceros they are arranged 1903.] NATURAL SCIENCES OF PHILADELPHIA. 391 in numerous oblique rows upon the surfaces of the coronal lobes. In Floscularia there is in most species a single row of cilia upon the lobes or on both the lobes and the interlobular coronal margin; in a few there are two rows upon the coronal margin {pe- lagica Rousselet, hoodii Hudson, cucullata Hood, trilohata Collins). All these cilia are preoral; they constitute, therefore, a trochus, and there is no postoral row (cingulum) on the external surface of the body behind the mouth. In fact, no postoral cihary ring occurs, for the diaphragm of Stephanoceros and Floscularia, wliich bears a row of cilia, and on that account was homologized by Hudson and Gosse with the cingulum of other Rotatoria, belongs to the intestinal tract and not to the coronal margin at all. In the young of Apsilus and Floscularia the mouth is ventral instead of terminal, as in the adult, and the anterior region of the ahmentary tract is cihated, these cilia not persisting to the adult stage. The diaphragm separates in Stephanoceros and Floscularia an anterior infundibulum from a poste- rior vestibulum, both without cilia; in Apsilus there is no diaphragm and no distinction of two chambers, and also no ciliation. In two species of Apsilus, in Floscularia and Stephanoceros an oesophageal tube is present. On the vestibulmn follows a non-ciliated proven- triculus, the posterior end of which is specialized as the mastax; next, a ciliated stomach; then a non-ciliated posterior intestine, then the short rectum opening into the cloaca. The intestinal ciliation is thus limited to the diaphragm, the stomach and (in some species of Flos- cularia) to the rectum. The foot is well developed in all but Apsilus; in this genus it is larger in the young than in the adult, so has probably degenerated. Only in F. chimccra Hudson does the foot terminate in two toes (it is doubtful whether this form belongs in the family). In all other forms it termi- nates in a single peduncle (except in Floscularia atrochoides Wier- zejski). The foot is strongly contractile, but not retractile into the trunk. The nephridial system is essentially alike in all three genera, likewise the nervous system; no ring nerves have been found in the coronal margin, and there is no suboesophageal ganglion. A dorsal and a pair of lateral coronal sense-organs are always present, and an additional pair of lateral ones in Apsilus. The germarium and oviduct are unpaired; the germarium is small and rounded in most species, but much elongated in Apsilus hucinedax (Forbes) and Floscularia ambigua Hudson. Distinct foot glands are absent in the adult; the only glands connected with the alimentary tract are one pair of stomach glands. A gland around the dorsal sense-organ is found in some forms. 392 PROCEEDINGS OF THE ACADEMY OF [May, The young are free-swimming, but it is the rule that the aduhs Uve permanently attached within a gelatinous tube. The exceptions are the pelagic species Floscularia pelagica Rousselet, atrochoides Wier- zejski, libera Zacharias. Since some of these swim carrying a gelati- nous tube, it is probable that they were derived from sedentary forms. All the known species are restricted to fresh water, except two which are eurhyaline. The sexes are dimorphic, and the small males represent individuals arrested in their development (they have considerable resemblance to immature females), and also degenerative in structure (shown by the absence of the alimentary tract). The whole structure is bilaterally symmetrical; even the unpaired dorsal sense-organ has a pair of nerves, which, as Zelinka has correctly argued elsewhere, points to its original paired condition. The radiality evinced by the arrangement of the coronal lobes is referable to the sedentary mode of Hfe. Finally, in contradiction to most other Rotatoria, in most species that have coronal ciha, these do not serve to create food currents, but are stiff and more like pseudopodia. Literature Cited. CuBiTT, C. Floscularia coronetta. Month. Micr. Journ., 2, 1869. Idem. Observations on Some Points in the Economy of Stephanoceros. Ibid., 3, 1870. Idem. On the Homological Position of the Tube-makers. Ibid., 8, 1872. DixoN-NuTTALL, F. R. On the Male of Stephanoceros Eichhornii. Journ. Roy. Micr. Sac, 1896. DoBiE, W. M. Description of Two New Species of Floscularia with Remarks, Ann. Mag. Nat. Hist., (2), 4, 1849. Idem. Stephanoceros Eichhornii. Ann. Anat. Physiol., 1, 1854. DuJARDiN. Histoire naturelle des Zoophytes. Infusoires. Suites a Buffon, 1841. Eckstein, K. Die Rotatorien der Umgegend von Giessen. Zeit. wiss. ZooL, 39, 1884. Ehrenberg, C. G. Ueber die Entwickelung und Lebensdauer der Infusions- thiere, etc. Abh. Akad. Wiss. Berlin, 1832. Idem. Dritter Beitrag zur Erkenntniss grosser Organisation in der Richtung des kleinsten Raumes. Ibid., 1835. Idem. Die Infusionsthierchen als vollkommene Organismen. Leipzig, 1838. EiCHHORN, J. C. Beytrilge zur Naturgeschichte der kleinsten Wasserthiere. Dan- zig, 1775. Forbes, S. A Remarkable New Rotifer (Cupel opagis bucinedax). Amer. Month. Micr. Journ., 3, 1882. FoxjLKE, S. G. On a New Species of Rotifer, of the Genus Apsilus. Proc. Acad. Nat. Sci. Philadelphia, 1884. Gast, R. Beitriige zur Kenntniss von Apsilus vorax (Leidy). Zeit. wiss. ZooL, 67, 1900. GoLDFUSS, G. A. Handbuch der Zoologie. Niirnberg, 1820. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 393 GossE, P. H. On the Structure, Functions and Homologies of the ^Manducatory Organs in the Class Rotifera. Phil. Trans. Roy. Soc. London, 1S55. Idem. The Crown Animalcule (Stephanoceros Eichhornii). Popular Science Review, 1, 1862. Idem. Contribution to the History of the Rotifera, 2. The Floscules, Ibid., 1862. . Grenacher, H. Einige Beobachtungen uber Raderthiere. Zeit. wiss. Zool., 19, 1869. Hood, J. Floscularia hoodi. Intern. Journ. Micr., (3), 5, 1895. Hudson and GossE. The Rotifera; or, Wheel-Animalcules. London, 1886-1889. Jennings, H. S. A List of the Rotatoria of the Great Lakes and of some of the Inland Lakes of Michigan. Bull. Mich. Fish Comm., No. 3, 1S94. Idem. Report on the Rotatoria. Ibid., No. 6, 1896. Idem. Rotatoria of the United States. Bull. U. S. Fish Comm., 1900. Leidy, J. Dictyophora vorax. Proc. Acad. Nat. Sci. Philadelphia, 1857. Idem. Rotifera without Rotatory Organs. Ibid., 1882. Idem. Dictyophora as Apsilus vorax. Ibid., 1884. Leydig, F. Ueber den Bau und die systematische Stellung der Raderthiere. Zeit. wiss. Zool., 6, 1854. Lord, J. E. A New Rotifer (Stephanoceros longispinatus?). The Naturalists' World, 2, 1885. LtTND, C. W. Danmarks Rotifera. Kobenhavn, 1899. ]\Iecznikow, E. Apsilus lentiformis, ein Raderthier. Zeit. wiss. Zool, 16, 1866. Montgomery, T. H. On Floscularia Conklin nov. spec, etc. Biolog. Bull., 1903. MoxoN, W. Notes on Some Points in the Anatomy of Rotatoria. Trans. Lin- nean Soc. London, 24, 1864. MtJLLER, O. F. Naturforscher. Halle, 1776. Oken, L. v. Lehrbuch der Naturgeschichte. 1815. Peirce, C. N. Remarks on Stepahnoceros Eichhornii. Proc. Acad. Nat. Sci. Philadelphia, 1875. Perty. Zur Kenntniss der kleinsten Lebensformen. Bern, 1852. Pritchard, a. a History of Infusoria, London, 1861. RosSETER, T. B. Tube of Stephanoceros Eichhornii. Science Gossip, 1881. Idem. Observations on the Life-History of Stephanoceros Eichhornii. Journ. Roy. Micr. Soc, 4, 1884. Stokes, A. C. Notes on the Genus Apsilus and other American Rotifera. Journ. Roy. Micr. Soc, 1896. Vallentin, R. Some Remarks on the Anatomy of Stephanoceros Eichhornii. Ann. Mag. Nat. Hist., (6), 5, 1890. Weber, E. F. Notes sur quelques Rotateurs des environs de Geneve. Arch. Biol, 8, 1888. Weisse, J. F. Verzeichnis von 155 in St. Petersburg beobachteten Infusorien- arten nebst Bemerkungen i'lber dieselben. Bull, math.-phys. Acad. St. Pet- ersburg, 3, 1845. Western, G. Notes on Rotifers, with description of 4 new species, and of the Male of Stephanoceros eichhornii. Journ. Quekett Micr. Club, 5, 1893. Explanation of Plates XVIII, XIX, XX, XXI. The figures, with very few exceptions, are freehand drawings, each con- structed from a considerable number of sketches or various details of structure. The degree of magnification is not uniform. These figures were made with the pen and were not reduced in size, which will account for occasional irregularities m the lines. The following is the significance of the abbreviations employed: A Sens.O., antero-lateral sense organ. ; Cer., cerebral ganglion. B.C., body cavity. ; Cir., cirrhus (penis). eel c C.2., contractor coronse primus CI., cloaca. ' and secundus. Cl.Ap., cloacal aperture. 394 PROCEEDINGS OF THE ACADEMY OF [May, Con.V., contractile vacuole. C.T., compressor trunci. Cut., cuticula. D.A., dorsal coronal arm or lobe. d.c, deflexor coronte. D.GL, dorsal gland. Dia., diaphragm. D.M., dorsal longitudinal visceral mus- cle. D.Sens.O., dorsal sense organ. d.tr. I-IV, depressor trunci primus- quartus. E., eye. F., foot. F.Gl., foot gland. Fl.C, nephridial flame cell. Ful., fulcrum inci. Hyp., hypodermis. Inf., infundibulum (coronal cup). L.A., lateral coronal arm or lobe. I.e., levator corona;. L.-C.M., longitudino-circular muscle. L.M., lateral visceral longitudinal mus- cle. L.Sens.O., lateral sense organ. M., mouth aperture. Man., manubrium mallei. Mast., mastax. Muse, muscles. Musc.C, muscle cell body. N., nucleus. N.c, nerve cell. N.com., nerve fiber commissure. Neph., nephridial canal. Neph.C, nephridial cell. Neph.com., nephridial commissure. N.f., nerve fiber. O., Ovum. Oes., oesophagus Oes.T., oesophageal tube. Ov., germarium. Ovd., oviduct. Ovg., ovogonium. Ped., peduncle of foot. P. Int., posterior intestine. Prov., proventriculus. R.c. I-VI, retractor coronse primus- sextus. Ram., ramus inci. Rec, rectum. R.M., retractor mastacis. Sph.An., sphincter ani. Sph.C. I-VIII, sphincter coronas prim- us-octaviis. Sph.tr. I-VI, sphincter trunci primus- sextus. Sp.S., sperm sac. Stoni., stomach. Stom.Gl., stomach gland. Test., testis. Tub., gelatinous tube. Un., uncus mallei. V.A., ventral coronal arm or lobe. Vac, vacuole. V.D., vas deferens. Vest., vestibulum. V.M., ventral visceral longitudinal muscle. X., cuticular ridge. Yk.N., necleus of yolk cell. Plate XVIII. — All the figures refer to Apsihis vorax. Fig. 1. — Dorsal view; this and figs. 2-4 represent fully expanded individuals. Figs. 2, 3. — Lateral (from the right side) and ventral views respectively, the details of the alimentary tract not slio^Ti. Fig. 4. — Newly hatched immature individual, about one-third the size of the adult, the musculature and cuticula not shown. Fig. 5. — Left half of the mastax viewed from above. Fig. 6. — Cerebral ganglion and its nerves, sense-organs, anterior portion of the nephridia, and contiguous muscles. Figs. 7, 8. — Two consecutive transverse sections of a cerebral ganglion, camera drawing. Plate XIX, Figs. 9-18. — Stephanoceros fimbriahis. Fig. 9. — Trunk seen from the left side, only the proximal portions of the coro- nal arms and the foot shown, details of the mastax not drawn. In this as in figs. 10 and 11 only the larger muscles are delineated, and those only in part. Fig. 10. — Dorsal view of trunk. Fig. 11. — Oblique ventral view of anterior trunk region with a portion of the coronal lobes. Figs. 12, 13. — The distal ends of two coronal arms, the first on dorsal and the second on lateral view. Fig. 14. — Distal end of foot with its peduncle. Fig. 1.5. — A nephridial flame cell. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 395 Fig. 16. — An entire individual enclosed within its gelatinous tube, ventral view, the larger longitudinal muscles sho'wn. Fig. 17. — Germarium, from a stained preparation. Fig. 18. — Mastax, postero-dorsal view. Figs. 19-21, Floscularia ambigua. Fig. 19. — Germarium seen from left side of the body. Fig. 20. — Dorsal view of anterior body region, musculature and corona cilia not shown. Fig. 21. — Immature individual pressed from the egg, seen from the right side; the musculature and mastax not drawn. Pl.\te XX, Figs. 22-24. — Floscularia coroneUa. Figs. 22, 23. — Lateral (from the left side) and dorsal views of anterior portion of the body; the cilia of the coronal lobes not dra-s\Ti, and only a portion of the musculature. Fig. 24. — Posterior end of foot with its peduncle. Figs. 25, 26. — Floscularia conklini. Fig. 25. — Anterior portion of body seen from the left side, mastax not drawn, and only a portion of the musculature. Fig. 26. — The entire animal with its gelatinous tube, from the right side. Figs. 27-35, Floscularia campanulata. Fig. 27. — Oblique ventral view of anterior trunk region, the nephridia not shown. Fig. 28. — A'iew from the left side; in this figure as in the preceding the coro- nal cilia are not shown. Fig. 29. — Germarium from a stained preparation, camera dra-wing. Fig. 30. — Germarium, oviduct and posterior portions of alimentary tract and nephridia of an immature individual. Fig. 31. — Antero-ventral view of the corona. Fig. 32. — Somewhat oblique lateral view of the cerebral ganglion and its nerves. Fig. 33. — A spermatozoon from body cavity of a female. Figs. 34, 35. — Oi^sophageal tube in two positions of movement. Pl.\te XXI. — AH the figures refer to Floscularia cam-panulata. Fig. 36. — A male from a tube of a female which contained 8 ova. This mature male measured not more than the length of the dorsal coronal lobe of the female. Viewed from the right side, the musculature and cuticula not drawii, from life. Fig. 37. — An entire individual within its gelatinous tube. Fig. 38. — Foot of an immature female, only about one-quarter the adult size, but attached. Fig. 39. — Mastax, dorsal view, muscles not drawn. Fig. 40. — Posterior end of foot and its peduncle of an adult individual. Fig. 41. — Nephridium of the right side of the body. 396 PROCEEDINGS OF THE ACADEMY OF [May, HISTORY OF THE CARIBBEAN ISLANDS FROM A PETROGRAPHIC POINT OF VIEW (Abstract). by per3if0r frazer, d.es-sc. Bibliography.^ Our knowledge of the geological history of the Antilles is still very imperfect. Among the titles constituting the bibliography of this subject are : Topography and Geology of Santo Domingo. Bv William M. Gabb. Trans. Am. Phil. Soc, XV, n. s. (1871). Observations and a Physico-Geological Description of the Regions of Hahana and Guanabacoa. Salterain, Madrid, 1880. A Pamphlet and Geological Map of Cuba, by Don Manuel Fernandez de Castro, based on work commenced in 1869. Madrid, 1881. On the Mountains of Eastern Cuba. W. O. Crosbj', 1882. Archean Character of Rocks of the Nucleal Ranges of the Antilles. Persifor Frazer. Bath Meeting B. A. A, S., 1888. Coleccion de apuntes sobre la riqueza minera de la prorincia de Santiago de Cuba. Published by Juan E. Ravelo. Sant. de Cuba, 1893. Reconstruction of the Antillean Continent. J. W. Spencer. Bull. G. S. A., August 14, 1894. Geographical Evolution of Cuba. Id. December 27, 1894. Zur Geologie von San Domingo. W. Bergt. Abhandlung der naturw. Gesell. "Isis" in Dresden, 1897. Cuba and Porto Rico, ivith the Other Islands of the West Indies. Robert T. Hill. The Century Co., New York, 1898. (This last work is a compendium of information on the subject and contains an extensive bibliography of the less scientific and more descriptive treatises on the West Indies.) At the Bath meeting of the British Association for the Advance- ment of Science in 1888 I presented numerous rock specimens and thin sections cut from them, illustrating a region of about forty miles around Santiago de Cuba. The rocks were partly eruptives and partly classic, but almost all exhibited profound alteration. The thin sec- tions from these eruptives were examined with me by Dr. Hensoldt, Mr. Kunz and Mr. La Croix in this country, and later by Mr. Teall, Mr. Rudler, the Abbe Renard, Prof. Judd and the lamented Prof. George H. Williams in London; all of whom were practically agreed as to the main constituents. The specimens were divided into : A. Those from the hills containing the West mine of the Jurugua ^ A more detailed statement of the observations, in the field and with the mi- croscope, on which these conclusions are based will be published later. — P. F. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 397 Iron Co., near Firmeza. (1) Diorites, some of which contained much altered hornblende and viridite (chlorite), the thin shdes filled with microlites and the rocks traversed by epidote veins. (2) Dolerites, diabase (gabbros), with chloritic ground mass, magnetite, rods of feld- spar and some olivine. B. From the hills southeast of that in which the "East mine" was located and about fifteen miles northeast of Santiago de Cuba. (1) Garnet rocks with iron ore (sp. gravity 3.962). (2) Fibrous actino- lite, and brown iron oxides partially altered to an epidotic mass. (3) Iron ores (some showing cross lines like the Widmanstatten figures in meteoric iron). C. From the Sietes Altares, about thirty-five miles east of Santiago de Cuba. Orthofelsite porphyry (rhyolites), like those erroneously referred to by the late Prof. H. D. Rogers as ''jasper," and later recognized by the late Dr. T. Sterry Hunt as a mixture to which he gave the general name " orthophyre " ; also like the Arvonian tuffs of Hicks, near St. David's Head, Pembrokeshire, Wales. D. The specimens from the region of the La Plata mines were quartz- ites containing hornblende, iron ores and, among the incidental min- erals, a claret-red garnet. In the area described were found upon or associated with the erup- tives sandstones, conglomerates and crystaUine limestones, laminated iron ores with masses of pyrites not yet converted into the latter. The alteration of the areas of contact in these rocks by the more recent diorite dykes which cut them was evident. From the character and relations of these rocks I deduced a physical continuity between the Archean of the mainland of the North American continent and the skeleton of the Cuban orographic system. From the zoological and geological researches of Alexander Agassiz in Caribbean and Mexican waters, and the careful studies by Gabb, Crosby, Spencer and Hill, the probability of very great changes of level in the Antilles since the close of the Cretacic period was estab- lished, and this probability is fortified by several different fines of proof, i.e., the ledges and shelves of the island borders, the wide distri- bution of the white radiolarian limestones, etc., and finally petro- graphic examinations of material from the several islands. Prof. Crosby pointed out orographic reasons for assuming a former " bridge " {i.e., causeway) between the Greater and Lesser Antilles. As he says, the mountains of "the northern arm of the island of San Do- mingo point toward Cape Maysi on Cuba," and the northern range in Cuba "regains the western trend and points directly toward Yucatan." 398 PROCEEDINGS OF THE ACADEMY OF [MaV, He also alludes in Ms paper of December 13, 1882, to the "axis of old eruptive rocks" of which, so far as he has been able to learn, " each mem- ber of the group consists." He does not give his authority for this statement nor say to what age he ascribes these eruptives; but if he contemplated the possibihty of its being pre-Cambrian he anticipated by six years two of the strongest grounds for my belief in the physical continuity of the Great and Little Antilles, and the present exposure of parts of the nucleus which are of great age and possibly have never been covered by sedimentary rocks. His observation that this nucleus is flanked on either side by schists and slates has been fully confirmed, and I have been tempted to class these with the mica schists of the Appalachians, and the feldspar por- phyry (rhyohtes) with the Arvonian tuffs of south Wales. Dr. W. Bergt strongly supports the Archean age of the nucleal axis of San Domingo if not of all the Caribbean Islands in the follow- ing words ■? "Das archaische Alter welches P. Frazer fiir die Centralketten des siidostlichen Cuba feststellen konnte, und das er fiir ganz Jamaika, fiir San Domingo, Puerto Rico, und die Windwardinseln vermuthete, kann nunmehr bestimmter fiir San Domingo angenommen werden." It may be, as Mr. Hill suggests, that no "Paleozoic nucleal rocks" have been established with certainty in "Cuba and Santo Domingo" or any other of the border lands of the "American ^Mediterranean,"^ although de Castro imagined he had discovered such near Cienfuegos, yet this fact would not preclude the possibility that part of these nucleal rocks are pre-Cambrian. To summarize the observations: (1) There is a complex of diorite, felsite and rhyolite forming the axes of the mountain ranges of eastern Cuba. (2) With these are associated schists which in places assume a gneissoid character. (3) The diorite and the clastic rocks resting on it are traversed by numerous veins of later eruptives, on the contact planes of which latter many phenomena of alteration may be observed. (4) No such indications of alteration are apparent at the contacts with the older diorite mass, but, on the contrary, the sedimentary rocks seem to have been deposited upon it without disturbance. ^ "Zur Geologic von San Domingo." Abh. der naturw. Gesel. "Isis" in Dresden, 1897, Heft II, p. 64. ^ Cuba and Porto Rico, p. 384. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 399 (5) All the rocks earlier than the intrusive veins, and even parts of these latter, show extensive alteration and probable transformation. One purpose of this abstract is to recall the fact that we have proofs of physical connection with the Western Continent of these outlying islands, not only from the physiographic features ; drowned valleys ; sub- merged plateaus; trend of conformation through the major axes of the present detached islands ; palseontological analogy with South American forms of hfe; etc., but, in addition to all these, the close petrographical relationship of the crystallized and crystalline rocks and their con- geners with those of the mainland. Note. A paper has just been received from Dr. Callaway (December 27, 1902) on the Plutonic complex of Central Anglesey,^ in which some very curious resemblances appear between the central complexes of Anglesey and Cuba, dissimilar as are the more recent mantles with which in the two cases these nuclei are clothed. In 1880 Dr. Callaway, in entire accord with Dr. Hicks and many of the non-official geologists of Great Britain as well as of our compatriot. Dr. T. Sterry Hunt, had established the pre-Cambrian age of these Anglesey rocks. In 1888 the Rev. J. F. Blake had confirmed a note previously made by Dr. Callaway of the production of schistosity in diorite by pressure. In the eastern district of Anglesey Dr. Callaway considers felsite and diorite the materials out of which the gneiss was made: "but ■s^'ith them is associated the well-known binary granite (haphte) originally called Dimetian by the late Dr. Hicks. A fourth variety .... is the quartz-felsite claimed by that geologist as Arvonian. It forms a part of the same magma as the granite, and must be carefully distinguished from c^uartzless felsite, into which it is some- times intruded in dykes and veins." After dividing the diorite and its modifications into hornblende-gneiss, decomposed diorite and chlorite-gneiss, micaceo-chloritic gneiss, and kersantite and biotite-gneiss, he says of the felsite that he "has never succeeded in obtaining a specimen in its original state," and in analysis it is almost indistinguishable from Vom Rath's rhyolite of the Euga- nean Hills, and from a rhyolitic obsidian from Medicine Lake. With the above are associated crystalhne limestones, of which the origin, according to Dr. Callaway, is chemical segregation and not sedimenta- tion. ^ Quart. Journ. Geol Soc, Vol. Iviii, 1902. 400 PROCEEDINGS OF THE ACADEMY OF [May The granite (haplite) and quartz-felsite, according to him, are in- truded into the diorite and felsite after the production of schistosity. Seams of quartz and feldspar (haphte) alternate with seams of biotite and feldspar (modified diorite) when the veins lie in the planes of shearing of the diorite, and the result is a banded gneiss. The quartz-felsite forms part of the same magma as the granite. Of two thin slides of the gneiss formed by schistose felsite intersected by numerous granite veins, Prof. Bonney "thinks much of the mica secondary." The paper concludes thus: " The following would seem to have been the sequence of events in the central complex. Diorite was first con- solidated. It was then penetrated by masses and veins of felsite, and blocks of it were isolated from the main mass (or masses) and floated off into the felsite. The consolidation of the felsite was the next stage. Earth-pressures then affected both diorite and felsite, producing scliistosity. A granite magma, usually haplite, sometimes quartz- felsite, then invaded the area, penetrating the diorite and the felsite in large masses, and sending into them countless veins, which commonly found their way along planes of schistosity, giving rise to banded gneisses." Summary of Results. "(1) The central complex of Anglesey was originally composed of diorite, felsite and granite. (2) The diorite has been modified into an elliptical dome of dark gneiss, namely, into simple gneisses by pressure, and into complex gneisses by pressure plus granitic intrusion. (3) The intrusion of the granite into the diorite has often produced fusion at the contact, sometimes with the generation of biotite in the diorite. (4) The diorite and dark gneiss form an insular mass surrounded by granite. (5) The felsite has been modified into quartzose and micaceous schists and gneisses by pressure, and into banded gneisses by the addition of granitic intrusions. (6) The quartz-felsites of the area are a part of the granitic magma. (7) Both diorite and felsite were modified into gneisses and schists prior to the intrusion of the granite and quartz-felsite, which are not foUated.". 1903.] NATURAL SCIENCES OF PHILADELPHIA. 401 POLYCHJETA FROM THE COASTAL SLOPE OF JAPAN AND FROM KAMCHATKA AND BERING SEA. EY J. PERCY MOORE. Hitherto our knowledge of Japanese Polychaeta has been hmited to collections made in the Httoral zone at several points along the south- eastern coast between Yokohama and Nagasaki and ably described in three papers by v. Marenzeller, a few^ species from the Japan Sea described by Grube and Mcintosh, and to the results of about half a dozen dredge and trawl hauls made by the ''Challenger" at points off Japan, which have been recorded by Mcintosh. At the close of an extended cruise in the South Pacific Ocean, the United States Fish Commission steamer "Albatross" spent the month of May and part of June, 1900, in dredging and trawHng along the con- tinental slope of Japan. During this cruise the vessel was in com- mand of Captain J. F. Moser, U. S. N., with Dr. H. F. Moore as natur- ahst, to wdiom, and to Dr. H. M. Smith, of the Fish Commission, I owe the pleasure of studying the rich and interesting collection of Polychseta taken.^ During May a line of about seventy dredging stations was run along the coast of Nippon or Hanshu from Yokohama westward about 200 miles, through Sagami and Suruga Bays and the Totomi Sea. These stations were mostly along the steep slope on the inner border of the Black Current, and about or within the 100-fathom hne, though the trawl was sometimes sent down to much greater depths. In early June, ten additional stations were estabhshed along the coast of north- ern Nippon, beginning at the lower end of Sendai Bay and extending for about 120 miles northward, entirely within the 100-fathom line. Later in June several dredgings were made in the shallow waters and muddy bottoms off Kamchatka, and others north of the Aleutian Islands in the southern portion of Bering Sea. The details of loca- tion, depth, character of bottom, etc., will be found in a Hst of dredg- ing records of the "Albatross" compiled by Mr. C. H. Townsend, and pubhshed in the Report of the U. S. Fish Commission for 1900. In this paper it has been thought necessary to give only the depth and 1 Acknowledgment is due to Hon. George M. Bowers, Commissioner of Fish- eries, for permission to publish the results in this form. 26 402 PROCEEDINGS OF THE ACADEMY OF [Juiie, general locality, together with a list of the station numbers at which each species was taken. All of the species contained in the collection have been determined and, with the exception of the families Sahellkhc and Serpulidce. and several species of other families which have been reserved for further study and comparison with material not now available, are recorded in the follo'^nng list. A large number of species, either new or hitherto unreported from that region, are added to the Japanese fauna. One of the surprises of this study has been the very small number of Maren- zeller's species which were taken by the "Albatross." Of the species previously known most of them belong to the circumboreal fauna, several were dredged by the "Challenger" at points off Japan, ajid the others have l^een recorded from various more or less distant parts of the world. As Marenzeller has in preparation an account of the geo- graphical relations of the Jaj^anese Polycha^ta, nothing more need be said on this subject. The general results remind one of those at- tained by ^'errill along the Gulf Stream slope of North America. POLYNOID^. Harmotlioe imbricata (Linn.) Mgrn, The setae of the two small specimens taken are colorless and the neuro- podials slender. Totomi Sea, 3,715, 65 fms.; Totomi Sea, 3,725, 12-13 fms. laenilla subfumida (Grube). This is one of the very few Philippine species which have been found on the coast of Japan, from which it has not been hitherto recorded. Sagami Bay, 3,702, 31-41 fms. Polynoa semierma sp. nov. (PI. XXIII, figs. 2, 3.) Form long and slender, composed of 64 somites and measuring 42 mm. in length and 3 mm. in breadth, exclusive of parapodia. Unfortunately the prostomium is in a very bad state of preservation and much altered, with all of the appendages, excepting one palpus, lost and only their position indicated by scars. The general shape of the prostomium is much as in Harmothoc, the anterior peaks being well- marked, pointed and widely separated. The left palpus has a length of four times the ^^•idth of the prostomimn and is slender and. regularly tapering. Two pairs of black, very small, circular eyes can be distin- guished, one lateral, about the middle, the other dorsal, near the pos- terior margin of the head. The protruded proboscis bears the usual jaws, but the papilla? have been destroyed. It has a length of 3.5 mm. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 403 The parapodia are long and slender and especially so toward the caudal extremity. Neuropodiiim conical, divided terminally into pre- and post-setal lobes, both prolonged dorsally, and the former the larger; ventral surface studded with large spherical papillae on short stalks. Notopodium a small, slender, freely projecting process supported by a distinct aciculum. Dorsal cirri with small ceratophores, and long, slender, regularly tapered styles which reach considerably beyond the setse tips anteriorly, but posteriorly, owing to shortening of the cirri and lengthening of the feet, fall short of the setse tips. Ventral cirri short, less than ^ dorsal, thick basalty and slender terminally; pos- teriorly they undergo the same relative change of position as the dorsal, and for the same reason. Only about i of the total number of elytra are in situ and are very easily detached, but the elytrophores indicate 25 pairs placed on the following somites: II. IV. V, then on alternate somites to XXIII, then on XXVI, XXIX, XXXII and XXXIV, and after that on every third somite to LXI. They are small, leaving a large portion of the back exposed, delicate and usually much distorted, nearly circular, with central attachment ' and quite without papillse or hairs of any kind; the medial half is reddish-brown, the lateral unpigmented. Notopodial setse are totally wanting, but there is a relatively stout tapering aciculum. The neuropodial setse are disposed in about 9 close rows; all are nearly colorless and rather slender. Those of the dorsalmost row or two form a distinct group which project beyond the others, and probably serve somewhat the purpose of the notopodial setse of other species ; their shafts are scarcely more than h the diameter of the other neuropodials and the ends are much prolonged and pro- vided with close transverse rows of fine hairs quite to the simple tips. The others have the usual form, the more dorsal ones with simple tips and as many as 14 transverse combs, the middle and ventral ones with a more or less prominent accessory tooth at the tip and shorter enlarged ends with as few as 8 combs on the most ventral setae. Sagami Bay, 3.698, 153 fms. Scalesetosus formosus sp. nov. (PL XXIII, figs. 4, 5. 6.) This s})ecies is referred to Mcintosh's genus on account of the form of the head and neuropodial setse, the latter being very characteristic ; in many other respects it departs widely from the generic type. The single example on which the description is based is complete, but broken into three pieces, and with the scales detached. At the poste- rior end is a small regenerating piece, but the form of that region, and especially the structure of the parapodia, indicates that it is of prac- 404 PROCEEDINGS OF THE ACADEMY OF [June, tically the normal length. The form is slender and tapers ver}^ sUghtly to the caudal end. The total length is 24 mm., the maxi- mum width of the body 2 mm., and between the tips of the parapodia 4.5 mm. Prostomium about as broad (across the ocular lobes) as long, marked for its entire length by a median dorsal groove wliich widens anteriorly to a broad and deep cleft separating the rounded frontal lobes or peaks ; the narrowest part of the prostomium is at the posterior border, ante- rior to which the lateral borders diverge, then swell out abruptly at about the middle as prominent rounded lobes which bear the anterior pair of eyes. Ceratophores of tentacles nearly spherical, possibly the result of contraction ; the median crowded between the cephahc peaks into the frontal cleft; the lateral partly beneath the median and well below the peaks ; style of median antenna lost ; the lateral very short and thick, scarcely longer than its ceratophore, and its diameter equal to ^ its length, ovate pyriform and little pointed. Palpi slender, tapered regularly to tip, about 2| times length of prostomium. Eyes 2 pairs, both very large, especially the anterior, but hghtly pigmented, the anterior pair situated on the ocular lobes, the posterior shghtly caudad and mesiad, and nearly or actually in contact mth the anterior. Tentacular cirri, or at least the ventral, which alone remain, about ^ as long as palpi, to which they have a similar form, with the distal half more slender. There are 57 somites in addition to those in regen- eration, and all but the peristomium and pygidium are setigerous. Ventral surface smooth, sole-hke, with rounded margins separated from the bases of the parapodia by a groove on each side. No visible nephridiopores or papillae. Elytrophores prominent, nearly cylindri- cal or shghtly tapering; the median space of the back scarcely exceed- ing in width their diameter. Parapodia large and prominent, similar throughout the series, except that the dorso-ventral diameter decreases toward the posterior end. NeuropocUum much larger than notopodium, broad, flat, leaf-hke, its lateral margin broadly rounded, the ventral convex, and dorsal con- cave, so that as a whole it appears to have a slight curvatm-e dorsad; at the end it is split into two vertical plates, both, especially the pos- terior one, being very thin and embracing the bases of the setae between them for a considerable distance. Notopodium a rounded lobe arising from the anterior side of the dorsum of the neuropodium about its middle. Like the el5^rophores, the dorsal cirrophores are remarkably large; although arising much nearer the base of the parapodia they reach as far laterad as, and at the same time much dorsad to, the noto- 1903.] NATURAL SCIENCES OF PHILADELPHIA. 405 podia ; styles formed like the palpi but more slender, reaching scarcely beyond the tip of the nem'opodia. Ventral cirri with short globular cirrophores placed about opposite to the notopodia; styles thickened basally but with filiform distal halves, and their tips just reaching the ventralmost neuropodial setre. First ventral cirrus closely approach- ing the tentacular cirri in length. Elytra 29 pairs, on II, IV, Y, then on alternating somites to XXI, then on XXII, XXIV, XX^'II, XXX. XXXI and again on alternating somites to LVII; all are in symmetrical pairs, and are easily detached, which is the condition of most of them. The first is nearly circular, the last somewhat triangular, the others broadly ovate with a sKght excentric posterolateral attachment. They are thin, smooth, color- less and translucent; with a very distinct nerve ramifying from the rear of attachment in dendritic fashion throughout the scale ; there are no cilia nor papillae, but numerous pellucid dots are scattered over the surface, in the center of each of which a small sense-organ appears. Notopodial setae few and irregularly arranged; they are short, reach- ing only half-way from their points of origin to the end of the neuro- podium, slender, rather strongly curved, pointed and with transverse rows of excessively fine teeth along the convex border and half way or more around the seta. Neuropodial setae arranged in a single vertical series which spreads in a broad fan-shaped figure, slightly separated into dorsal and ventral halves, slender and very long, probably quite equalling the parapodium when fully protruded, somewhat enlarged and bent below the slender, tapering, very finely serrulate, slightly hooked and undivided end ; at the region of the thickening is a half ring of long, fine, comb-like teeth supported on a slight shoulder, which is placed on the ventral side of the setae in the dorsal half of the bundle and the dorsal side of those in the ventral half. The most ventral setae have shorter, stouter and more strongly hooked tips. Both notopodial and neuropo- dial setae are colorless and beautifully transparent. The setae of the first setigerous foot differs in no appreciable respect from the others. Only the type knoTvm from Sta. 3,703, Sagami Bay, 31 fms. Lepidonotus cMtoniforinis sp- nov. (PI. XXIII, figs. 10, 11.) Form short and broad, with a regular elliptical outline, depressed; the greatest width, whether measured between the tips of the setae or the margins of the body, at the middle. The type measures 37 mm. long, 16 mm. wide, and 6 mm. deep. Prostomium roughly square, the anterior and lateral margins slightly convex, the posterior concave; a pair of prominent posterior lateral ocular lobes, and the production of the anterior face into the 406 PKOCEEDINGS OF THE ACADEMY OF [June, lateral tentacular ceratophores modify this form. Eyes 2 pairs; the anterior large and partly on the base of the posterior lateral lobes, as a consequence of which they face both forward and outward ; the pos- terior are about ^ the diameter of the anterior and are situated nearly on the dorsal surface close to, but slightly mesiad and caudad of, the ocular lobes. Ceratophore of median tentacle about f as long as prostomium, stout and distally swollen; style 4 times length of prostomium, slender, slightly tapering, with an abrupt egg-shaped subterminal enlargement having the large end distad, and followed by a terminal filament of equal or shghtly greater length. Lateral tentacles similar in form to the median but only about f as long, and the ceratophore only ^ as long as that of the former. The colors of the tentacles are well pre- served and striking in their contrast with the colorless head; the cer- atophores are ringed with yellowish-brown, the basal half of the style strongly tinged \\ath brown, the proximal half of the subterminal en- largement deep brown, and the rest pure white. Palpi about equal- ling median tentacle, rather stout, but their bases not enlarged nor extending laterally beyond prostomium, gently tapering, abruptly sharp pointed, but without terminal filament, densely ciliated, basal half pale brown. Peristomial parapodium long, slender, reaching to the level of the tip of the tentacular ceratophore. Tentacular cirri similar in form and color to the tentacles but \vith somew^hat longer terminal filaments, the dorsal equalling the median tentacle, the ventral somewhat shorter. There are 26 somites, of w^hich 25 bear setse and 12 elytra. The ne- phridial tubercles occupy the usual position and point strongly laterad and slightly ventrad and caudad ; the last pair, on XXVI, are in con- tact mesially and all except the first, on VI, are prominent. Prominent dorsal papillae occur on all the somites from II to XX in- clusive, being median on II and on X^' to XX, and in double parame- dian series on III to XIV inclusive. The first one is a low, flat, nearly c{uadrate nuchal tubercle which somewhat overlaps the prostomium from behind ; from III to XIV each somite bears a larger anterior and a smaller posterior pair; on III those of each pair are close together and united by slight transverse ridges; the others are well separated, the median interspace increasing in width with the segments ; somites XV to XIX have each a single larger anterior and a smaller posterior papilla, and XX the large one alone. The elytrophores are large and shaped like the body of an oyster or even more like the human external ear viewed from the cephalic side. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 407 On the cirri-bearing somites are smaller, and structurally somewhat simpler, but essentially similar structures, through which transverse muscles pass into the parapodia and dorsal cirri. As will be indicated below, they bear branchial filaments similarly to the ely trophores, with which they are obviously homologous. Branchial filaments occur on every somite from III to XXIV inclu- sive, except XXIII, the last somite bearing elytra, and minute rudi- ments on the anterior side even of XXV and XXVI. They appear to be mere hollow integumental sacs generally of simple finger-hke form. On elytra-bearing somites they are disposed as follows : One arises from the antero-external and one from the postero-external margin of the elytrophore. The former is present on every elytrophore except the first antl last, is unbranched but bends sharpl}^ laterad at a right angle. The latter is longer, and on typical somites divides into a short medial and a long lateral branch diverging at right angles from the short stem ; it is altogether absent on the first and the medial branch is wanting on the 11th and a few of the anterior ely trophores. On the parapodium are 2 dorsal, 5 or 6 anterior and 4 or 5 posterior filaments (in addition to 1 or 2 rudimentary ones on the base of the ely trophores), the most ectal one in each case being bifid. Anterior to the 5th and posterior to the 11th elytra-bearing somite this arrangement is simplified by reduction in the number of filaments. Essentially the same arrangement obtains on the cirri- bearing somites, but the filaments are more numerous, larger and sometimes even trifid. The elytrophore filaments are represented anteriorly by 2 simple ones, and posteriorly by 3, of which the middle one is bifid. There are usually 5 dorsal parapodial filaments, of which one at the base of the cirrus, like the most external on the anterior and' posterior faces of the parapodia, is usually trifid. The parapodia are stout, with the neuropodium of a somewhat com- pressed, obliriuely truncated conical form and the notopodium a rather prominent short tubercle on the anterior dorsal face near the base of the parapodium. The ventral surface of the parapodia, and indeed of the entire body, is covered with a close nap of short, fine processes. Ventral cirri with low tubercle-like ceratophores borne on the middle of the ventral neuropodial surface; their styles short, scarcely reaching half-way to the end of the foot, the basal half stout and swollen, the distal tapered to a sharp point, but with no proper filament. The dorsal cirri are quite long, their filamentous tips at least reaching quite beyond the setse line; they have exactly the shape and color of the tentacular cirri, but on the base of each is a prominent bilobed glandu- 408 PROCEEDINGS OF THE ACADEMY OF [Juiie, lar swelling; the last pair, instead of being longer, is shorter than the others; otherwise both ventral and dorsal cirri present the usual modifications toward the end of the body. The very regular outUne of the body, when seen from above, is due to the dense tufts of dorsal setae which project beyond the margins of the scales and conceal all but the tips of the ventral setse, and from l^elow to the remarkably even arrangement of the neuropodia and the ^Tntral setse. The dorsal setse are very numerous and spread chiefly in a horizontal plane, but in such a way as to overlap successively from behind forward. They are very delicate, capillary, tapering from base to tip and furnished with fine but distinct opposite or nearly opposite processes of a length 3 to 4 times the diameter of the stem. The ventral setse are arranged in very regular horizontal rows, 7 or 8 of which are subacicular and 2 supraacicular ; except the dorsalmost and ventralmost rows, which have fewer, each contains 4 setse. They are of a beautiful pale amber color, and transparent, moderately stout, shghtly curved and with the tip smooth and strongly hooked. As in Euphione they are densely bearded toward the end, about -g- or j, according to the position of the seta, of the exposed part being provided with numerous rows of long fine hairs, the terminal ones of which curve backward and envelop the tips. On the first setigerous somite the ventral setse are slender and tapering, with a close beard of fine hairs. Twelve pairs of elytra are borne on somites II, IV, \. ^ II, IX. XI, XIII, XV, XVII, XIX, XXI, and XXIII. They are large, strongly imbricate and decussate, and so close-fitting that their boundaries can scarcely be distinguished except at the margins; the attachment to the elytrophores is unusually firm. The first is irregularl}' circular with a nearly central scar, the last roughly triangular, and the others more or less bean-shaped, the more anterior ones being deeply emar- ginated and asymmetrical. The 7th is the largest and the size falls off each way to the 1st and 11th, the 12th again becoming larger. Cilia form a strong marginal fringe around more than f of the circum- ference of the scale, only the covered anterior portion being free ; they are longest and coarsest on the lateral half of the anterior border, from which large papillse are wanting, and where they are about equal to the longest papillae present on the scale ; a small group of rather large ones occurs also on the medial margin. Small cilia are scattered spar- ingly over the greater part of the exposed surface of the scales and are larger and more numerous on the area external to the scar ridge. The papillae are remarkable for their size and form, and cover the entire upper surface. The anterior medial region and a narrow area along the ante- 1903.] NATURAL SCIENCES OF PHILADELPHIA. 409 rior lateral border bear only small, low, smooth papillae; from these areas they gradually differentiate into several distinct forms. Lower flat papillae with from 4 to 8 peripheral radiating spines occur chiefly on the postero-lateral half ; a few especially large examples of this type which lie just behind the scar ridge may bear 1 or 2 apical spines in addition to the peripheral ones. Smooth globular papillae without spines are occasionally found with the latter. Along the scar ridge are 1 to 3 (of which the mesial one is the largest and most character- istic) very large papillae having the form of an irregular inverted cone, the enlarged end of which is thickh' studded with rough processes. On the entire area between the lateral end of the scar ridge and the postero-lateral margin are numerous large, pale or colorless, erect, clavate papillae with slender pedicels, and studded with peculiar rough scaly nubs. A single marginal rank of similar but larger, recumbent and usually brown papillae extends around the entire postero-internal border. The papillae vary much in color; those on the ridge are gen- erally yellowish-browm, the others varying shades of yellowish-gray, gray with yellow spines, brown or nearly l^lack. The darkest are usually found in a small group just mesiad of the scar ridge. The 1st scale bears a central group of the large rough papillae and a complete marginal circle of clavate ones, beneath which is a circle of short cilia. The last has club-shaped papillae on its lateral margin only, and nearly its internal half is free from cilia. When the scales are in place the entire animal presents a striking superficial resemblance to a Chiton. The rough scale ridges con- verge in each pair to form a series of \'s pointing forward. T}T3e, Sagami Bay, 3,700,63 fms.; Totomi Sea, 3.733, 49 fms.; also one from an unknown station. Lepidonotus branchiferus sp. nov. (PI. XXIII, figs. 7. 8, 9.) Outline short, broad, elliptical, very regular, somewhat depressed. Length 26.5 mm., width to tips of setse 14 mm., to margins of scales 11 mm., depth 6 mm. Prostomium decidedly wider than long, its lateral margins wdth prominent preocular protuberances, making this the broadest region; eyes two pairs, rounded, black, the anterior much the larger and more prominent, the posterior distant from them less than their own diam- eter mesiad and caudad. Median tentacle with stout cjdindrical basal piece slightly shorter than prostomium ; style about 4 times the length of prostomium, tapering to a slender region beyond which is a subter- minal ball and a short terminal filament. Lateral tentacles with ceratophores less than ^ that of median tentacle, and continuous with 410 PROCEEDINGS OF THE ACADEMY OF [Juiie, frontal processes of head, which are shghtly dorsad of the level of the median ceratophore; style similar in form to median bnt with a rela- tively longer terminal filament which reaches only to the median ball. Palpi with very broad, widely separated bases, fully half of which pro- ject beyond the sides of the head ; otherwise they are similar in form to the palpi of L. cJiitoniformis; basal -f brown, the rest white. Peristomial parapodia not cpiite reaching to tip of median tentacular ceratophore; the dorsal tentacular cirrus, which alone remains in the only specimen, similar in form to median tentacle, but with a longer filament, and the entire style slightly shorter. The protruded probos- cis has a length of 6 mm, and a diameter of 3.7 mm. It bears four light brown fang-like jaws of the usual form, and thirteen papillae above and thirteen below, all but the small lateral ones, which are simple, being strongly curved, compressed and bilobate. There is also a trans- versely elongated low subterminal ventral papilla*. Total number of somites, including peristomium, 26; setigerous somites 25; elytra 12 pairs. Nephridial pai^illae may be detected as far forward as IV, but the first 3 are very small, the others exactly as described for chitoniformis. Except that the difference in size between the anterior and posterior ones of the dorsal series is less evident, the dorsal tubercles are exactly as in L. chitonijormis. No important differences are to be noted in the form and structure of the elytrophores. The branchiae have the same arrangement as in L. chitoniformis and, although fewer and simpler, are larger and more conspicuous than in that species. All are rather long and slender and entirely unbranched, and are directed more or less latei'ad and dorsad. None occurs on the dorsum of the parapodium. On the elytra-bearing somites one arises from each the posterior and anterior Ijorder of the elytrophore, the former being somewhat the larger and more laterad. On the anterior face of the parapodium is a series of about 3, arranged along a line from the one mentioned above to the base of the ventral cirrus, to which, however, it does not reach ; one or two smaller detached processes may lie ventrad of the middle of this series. On the posterior face usually 3 in a group are found below the outer end of the elytrophore, at a lower level than the anterior series. On the cirrhiphorous somites the arrangement differs slightly. A short anterior one arises from the border of the ridge passing to the cirrus ; an oblique series of 4 occurs on the anterior face of the parapodium. Posteriorly there are two mar- ginal, a cirral which arises from the base of the cirrus in actual contact with its posterior gland, and 2 or 3 others lower down on the posterior 1903.] NATURAL SCIE^XES OF PHILADELPHIA. 411 face of the parapf)diuin. Toward the ends of the body the number of branchial filaments diminishes. The parapodial dorsal and ventral cirri are much the same as in L. chitinoformis , but the dorsal cirrus is somewhat shorter, just barely reaching the ventral setse tips; its knob is generally more spherical, and its basal glands, while similarly arranged, much longer. Both dorsal and ventral setse are similar in their arrangement to those of L. chitoniformis . The former are shorter than in that species, leaving the entire bearded portion of the ventral setse exposed from above, and are more slender, softer and more densely provided with longer lateral processes, which are arranged either in tufts or whorls, the exact arrangement being uncertain owing to the presence of an incrustation of foreign matter. The latter are somewhat stouter, with shaggier beards of fewer rows of hairs; one was found with a terminal sheath still partly in place. In number, arrangement, form, and even in the character of their papillae and cilia, the elytra of this species approximate those of L. chitoniformis. The chief difference lies in the larger size of the granules and the strong tendency of the papillae to become spinous. The papillae immediately surrounding the low scar ridge are of a stellate form with long, sharp-pointed rays bearing one or more conical spinous on the upper side. The largest of this tj^pe are posterior to the ridge, where occur also a few nearly globular papillae bearing a single long conical spine. On the medial side of the scar ridge these papillae become gradually more irregular and oblique, finally passing into the smaller papillae whose pointed summits are directed toward the pos- terior margin of the scales. Toward the covered area of the scale all papillae become smaller and smoother, and soon low, rounded and color- less. Clavate papillae occur in the lateral area, and in a postero-lateral marginal row ; they are relatively smaller and much more slender than in L. chitoniformis and bear numerous prominent conical points instead of rough nubs. The erect ones of the lateral area bear a number of spines, varying with the size of the papilla, along all parts of their sides and summits, whereas the reclining ones of the marginal row have spines on the upper or exposed surface only. Over the region of the scar, in addition to the stellate papilke, which there attain their maxi- mum size, are from 4 to 6 particularly prominent papillae crowned with numerous short spines arranged in a tuft. In spite of their rough- ness all of the papillae, even the largest, are soft. The covei'ed i3ortions of the scales and the lateral zone are colorless, the middle and pos- terior parts yellowish-brown, against which the various colored papilkifi 412 PROCEEDINGS OF THE ACADEMY OF [JlUie, stand out clearly. The larger papillae are chocolate-brown, recldish- brown, yellowish-brown, partly brown and partly yellow, or, in strik- ing contrast, white, and it is very seldom that two adjacent ones are of the same color. One specimen, type, Sta. 3,702, Sagami Bay, 31-43 fms. L. branchiferus and L. chitoniformis are evidently closely related to L. giganteus Ivirk from New Zealand which, according to Thomson, possesses essentially similar branchiae and dorsal tubercles, and exhibits additional features of resemblance in the elytra and sette. In a number of respects all three depart from the typical species of Lepidonotus and might very properly be segregated as a distinct generic group. Lepi- donotus branchiata (Treadwell) from Porto Rico possesses similar branchiae, but the setae and elytra are quite different and no reference is made in the description to dorsal tubercles. Lepidonotus caelorus sp. nov. (PI. XXIII. fig. 12.) The form is, as usual in the genus, short and compact, but much more slender than the 2 species just described; the largest specimen measures 25 mm. long, exclusive of the tentacles and anal cirri, and 8 mm. wide to the tips of the setae. Number of somites 26. 25 bearing setigerous parapodia, and 12 elytra. Ventral surface smooth, with nephridial papillae from ^TII to XXV inclusive; all but the first are prominent and tubular and project freely caudo-laterad. Prostomium slightly wider than long, though the continuation of the peaks into the bases of the lateral tentacles gives it a somewhat elongated aspect. Eyes normall}' 2 pairs, which have coalesced in the type; the posterior near the caudal end of the lateral surface, but not concealed by the nuchal fold, looking latero-dorso-caudad, heavy black, circular; the anterior lateral, at the point of greatest width, looking latero-dorsad, larger than the posterior and elliptical or crescen- tric, black. Except the palpi the cephahc appendages are very easily displaced, and the median tentacle is present in the type alone. Its total length is about 5 times the head, of which ^ is made up of the basal piece and yV of the filamentous tip; l^asal piece constricted at the middle, the proximal portion more opaque and smoother; style slender throughout and tapering to a slightly subterminal enlargement, beyond which it diminishes suddenly to the filiform tip. The lateral tentacles are more often preserved and vary considerably in length. They have the same general form as the median tentacle, but are longer and decidedly more slender, with the subterminal enlargement scarcely evi- dent, and the filiform tip longer (up to ^ of the entire length) ; the basal 1903.] NATURAL SCIENCES OF PHILADELPHIA. 413 piece, which is continuous with the prostomial lobes, is shghtly shorter, and arises at a shghtly more ventral level. Palpi about 3 times length of prostomium, tliickened basally, tapered to end, which bears a very short terminal filament. The peristomium presents no noteworthy features. Tentacular cirri similar to lateral tentacles, the dorsal somewnvhat exceeding median tentacle, the ventral slightly shorter; the slender parapodium, which supports them, reaching beyond the tentacular basal pieces. The typical parapodium (X) presents the following features: neuro- podium large, nearly truncate or slightly angulated at the point from which the deep brown aciculum protrudes, ventral margin horizontal, dorsal sloping with a slight curve to the elytrophore ; notopodium amere lobe on antero-dorsal face of neuropodium, supported by a slender aciculum. A'entral cirri with a short filiform appendage, the tip of which falls short of the bases of the neuropodial setce ; dorsal cirri borne on prominent ceratophores which have a dorso-caudad position with relation to the foot, similar in form to the tentacular cirri, about 2^ times as long as the ventral cirri, of which length the ceratophore constitutes -J. The ventral cirrus of the 1st foot is, as usual, longer; on the last two the dorsal ceratophores become posterior; the last foot lacks the noto- podium, or at least the notopodial setae. The anal cirri are the longest appendages of the body, fully twice the dorsal cirri, and bear very long filiform tips. Twelve pairs of elytra are borne, on somites II, III, IV, VI, VHI, X, XII, XIV, XVI, XMII, XXI and XXI^^ They are strongly imbricated and tough, membranous and firmly attached. With the exception of the first and last they are elongate-pyriform, or more ovate posteriorly, attached posterior to the middle, and with the slender pre-peduncular portion covered by the preceding scale; the first is broadly ellipsoidal, the last roughly triangular with the longest side mediad and the angles rounded. With the exception of a narrow smooth area at the anterior end, the entire dorsal surface of the scale is thickly clothed with papillae of various kinds. Anteriorly are a few small conical spines with apices directed obliquely toward the posterior margin of the scale. Farther back they become larger and more numerous and distinctly differenti- ated into two forms distributed to the internal and external halves of the scale. On the external area they remain smaller and depart less from the simple conical form, but most of them develop a few jagged points at the apex and become more elevated, especially those of a marginal series, which are larger and somewhat clulj-shaped. On 414 PROCEEDINGS OF THE ACADEMY OF [Juiie, the internal area a gradual transition into larger, rounded, rough papillae takes place, especially over the area of attachment where this kind occiu's nearly exclusively; but elsewhere they are interspersed with papilla3 of the same type as, but larger than, those of the external area. The large papillce are of a globular or haycock form and reach a diameter of 10 or even 20 times that of the spines. Their surface markings are very peculiar and characteristic, somewhat resembling the chasing of a cane or umbrella head, but rougher than such work is customarily. >Sometimes the markings are very regularly arranged in rows converging to the apex, and may then be simply roughened ridges and grooves, or rows of overlapping scales or even spines. The first and last elytra are the roughest of all and have the largest papillae. A strong fringe of long cilia marks the posterior external margin of typical scales and nearly encircles (except for a small part of the inter- nal margin) the first. The longest have a length of about ^ or | of the greatest transverse diameter of the scales, but become much shorter on approaching the mesial side, along the wdiole exposed })ortion of which they are continued as integumental sense-organs of gradually diminishing length. The notopodial setae form a large spreading tuft, but their tii:)S scarcely reach beyond the end of the neuropodium ; they are pale hay color, capillary, bipinnate, with the lateral processes alternating. Owing to a constant coating of foreign substances few details can ])e made out. Neuropodial sette arranged in 3 supraacicular and 5 sub- acicular horizontal row«, amber-colored, relatively slender, with the smooth tips unusually long, and except on the most dorsal, exceeding the spinous portion in length ; 4 transverse rows of spines on the ven- tralmost setae, 9 on the dorsalmost. Some color is retained in the elytra, which are yellow or l^rown, some- times with an irregular greenish blotch surrounding a pale area opposite the point of attachment, the larger papillae usually dark brown. The dorsal cirri and all cephalic appendages, with the exception of the palpi, have a broad zone of dark green pigment above the base, and a deeper but narrower one just proximad of the subterminal enlargement. LcpkJonotus caiorus somewhat resembles L. pleiolepis von Marenzel- ler, but differs especially in the numerous spheroidal papillae on the elytra and the more slender form and longer smooth tips of the neiu'o- poclial setae. Sagami Bay, 3,698, 153 fms., type and 8 other specimens; Suruga Bay, 3,707, 63-75 fms. 1903.] NATURAL SCIENCES OF PHILADELPHIA. 415 Xepidonotus (Hylosynda) vexillarius sp. nov. (PI. XXIII. fl<;s. 1:5. 14, 13.) A rather slender species, measuring 23.5 mm. in length. 2.5 mm. in maximum breadth of body on the ventral surface of X. 4 mm. to ends of parapodia, and 5.5 to tips of setse. Prostomiimi slightly wider than long, sides smoothly convex without any prominent lateral lobes, posterior margin for its middle half con- tinuous with peristomium. frontal sinus very shallow. Eyes 2 pairs, black, circular; the anterior larger and situated close to lateral margins at widest part of head; posterior on postero-lateral curvature, looking ■caudad, dorsad and laterad. ^ledian tentacle arises from dorsal sinu