PROCEEDINGS ^O^^O J 20S I M U2 - 09 Biological PRINTJ 1898-95. PROCEEDINGS OF THE Biological Society of Washington VOLUME VIII. 1893. n WASHINGTON : PRINTED FOR THE SOCIETY. 1898-95. COMMITTEE ON PUBLICATIONS. F. H. KNOWLTON, Chairman. T. H. BEAN. L. O. HOWARD. F. V. COVILLE. T. S. PALMER. ^"''^ ^\\\^[.\> %/ - 1 1 a T r -r ' TERRY BROS., Printers. CONTENTS. PAGE Officers and Committees for 1893 . . . . iv Joint Commission for 1893 . . . . . iv Proceedings ........ v-xiii Addresses and Communications : A Jumping Mouse (Zapus insignis Miller) new to the United States. By Gerrit S. Miller, Jr., 1 Paleontology of the Cretaceous Formations of Texas. The In- vertebrate Paleontology of the Trinity Division. By Robert T. Hill, 9 Further notes on Yucca Insects and Yucca Pollination. By C. V. Riley, 41 Description of a new White-footed Mouse from the Eastern United States. By Gerrit S. Miller, Jr. , 55 Development of the Crachial Supports in Dielasma and Zygos- pira. By Charles E. Beecher and Charles Schuchert. And on the Development of the Shell ot Zygospira recurvirostra. By Charles Schuchert, 71 Re-discovery of the Mexican Kangaroo Rat, Dipodomys Phil- lipsi Gray. By C. Hart Merriam, M. D. With field notes by E. W. Nelson, 83 The Paleontology of the Cretaceous Formations of Texas. The Invertebrate Fossils of the Caprina Limestone Beds. By Robert T. Hill, 97 Two new Wool- rats from the plateau region of Arizona (Neo- toma pinctorurn and N. Arizonse), with remarks on the Validity of Genus Teonoma of Gray. By C. Hart Merriam, M.D., 109 Notes on Thomomys bulbivorus. By Gerrit S. Miller, Jr., . . 113 Description of a new Species of Lagomys from Alaska. By E. W. Nelson, 117 Juncus rnarginatus and its Varieties. By Frederick Vernon Coville, 121 Description of eight new Ground Squirrels of the genera Sper- mophilus and Tamias from California, Texas and Mexico. By C. Hart Merriam, M. D., 129 Description of a new Species of Arvicola of the Mynomes group from Alaska. By E. W. Nelson, 139 Preliminary Descriptions of four new Mammals from Southern Mexico, collected by E.W. Nelson. By C.Hart Merriam, M.D., . 143 The Yellow Bear of Louisiana, Ursus luteolus Griffith. By C. Hart Merriam, M. D., 147 Index, ... 153 LIST OF THE OFFICERS AND COUNCIL OF/THE BIOLOGICAL SOCIETY OF WASHINGTON. ELECTED DECEMBER 31, 1892. OFFICERS. PRESIDENT. C. V. RILEY. VICE-PRESIDENTS. FRANK BAKER, RICHARD RATHBUN, B. E. FERNOW, C. D. WALCOTT. RECORDING SECRETARY. F. V. COVILLE. CORRESPONDING SECRETARY. F. A. LUCAS. TREASURER. FRANK H. KNOWLTON. COUNCIL. TARLETON H. BEAN, C. HART MERRIAM,* WILLIAM H. DALL,* T. S. PALMER, THEODORE GILL,* THEOBALD SMITH, G. BROWN GOODE,* FREDERICK W. TRUE, L. O. HOWARD, LESTER F. WARD.* STANDING COMMITTEES 1893. Committee on Communications. B. E. FERNOW, Chairman, CHARLES SCHUCHERT, ERWIN F. SMITH. Committee on Publications. THEODORE GILL, Chairman. L. O. HOWARD, T. W. ST ANTON, FRANK H. KNOWLTON, GEO. B. SUDWORTH. Committee on Trees and Shrubs. LESTER F. WARD, Chairman. F. V. COVILLE, THEODORE HOLM, FRANK H. KNOWLTON. Delegates to the Joint Commission of Scientific Societies of Washington. RICHARD RATHBUN, C. V. RILEY, LESTER F. WARD. *Hx-Presidents of the Society. Two HUNDRED AND FOURTH MEETING, January U, 1893. Vice-President Dr. Baker in the chair, and thirty-one mem- bers present. Mr. Walter H. Evans was elected an active member. Dr. Erwin F. Smith presented a paper entitled ADDITIONAL NOTES ON PEACH ROSETTE.* Mr. D. G. Fairchild presented some NOTES ON APPLE AND PEAR FUSICLADII. The paper was discussed by Messrs. Erwin F. Smith and Waite. Prof. L. F. Ward spoke upon THE NEW BOTANY, f Prof. Ward's communication was discussed by Dr. Baker. Drs. Theobald Smith and V. A. Moore followed with a com- munication upon THE GROWTH OF BACTERIA THROUGH THE PASTEUR- CHAMBERLAND FILTER.;}: Two HUNDRED AND FIFTH MEETING, January 28, 1893. Mr. L. O. Howard in the chair, and thirty-one persons present. Messrs. D. D. Luke and H. H. Doubleday were elected active members. Prof. Robert T. Hill presented some NOTES ON THE PALE- ONTOLOGY OF THE COMANCHE SERIES. The paper was dis- cussed by Messrs. Stanton and Ward. * Journal of Mycology, vii, pp. 226-232, 1893. tScience, N. Y., xxi, pp. 43-44, Jan. 27. JZur Priifung der Pasteur-Cham berl and Filter, rozoon in the Intestin al Villi of Cattle. s 1 Head and body. Tail vertebrae. tf H3 a S Ear from notch. Skin. Skull. 787 1414 1836 1840 1841 1842 1846 583 584 585 586 587 588 620 622 623 624 625 626 627 628 630 631 .1622 1635 1636 1646 1650 1663 1669 1679 1703 1704 1705 1427 1944 1945 1946 1947 1948 Oak Bay N B. Sept. 19, '91 Oct. 7, '91 "9" 9 | 9 | 9 9 9 | 9 | c? 241 190 215 203 215 215 225 217 198 217 214 221 216 208 185 217 215 231 214 219 194 209 200 215 205 200 202 203 203 198 198 229 205 208 148 114 134 132 138 135 140 134 143 132 135 142 139 128 116 127 137 152 129 132 122 128 122 124 131 128 120 125 124 121 118 129 120 125 128 123.5 131 148 117 137 33 28 31.4 31 30.6 31 32 30 31.2 30 31.6 31.5 30 29.6 29.5 28.8 30.4 31.4 30 30.6 27.5 29 28 31 30.4 30 29 28 28 28 29.6 30.8 29.6 30 29 28 29.5 29.5 27 30.5 13* 11 12.4 11 11.6 11 11 12 12.8 11.8 13 12.8 13.4 12.2 10.5 13 12.4 13 12 11.4 12 11.4 11 14.6 13 12.8 13.8 13 13 12.8 13.8 14.4 14 14 10f 10. 1 8.5 12.5 10 14 'leii' 1618 1619 1620 1624 517 518 519 520 521 522 538 539 540 541 542 543 544 545 ' 546 547 1419 1432 1433 1442 1446 1459 1464 1474 1495 1496 'i239 1718 1719 1720 1721 1722 U I it it I ti t it t Peterboro, Madison Co., N. Y. tt a a it a it it tt tt tt a a a a n tt a a a a it tt it it tt a a tt it tt tt a tt tt < a it ti it it ti a a a ' a a AVareham,Plymouth Co., Mass. a it it n a n a a it a July'i7,'9l' ' 17, '91 ' 17, '91 ' 17, '91 ' 17, '91 1 17, '91 ' 20, '91 1 21, '91 ' 21, '91 ' 21, '91 ' 21, '91 ' 21, '91 ' 21, '91 ' 21, '91 ' 21, '91 ' 21, '91 ' 30, '92 Aug. 1,'92 ' 1, '92 ' 9, '92 ' 18, '92 ' 25, '92 ' 26, '92 Sept. 7, '92 ' 17, '92 ' 17, '92 ' 17, '92 May 28, '92 July 13, '92 Aug. 18, '92 July 7, '92 " 13, '92 Aug. 12, '92 88 ' ' 65 89 86 63.5 83 * Collected by H. H. Me Adam. f Collected by Outram Bangs. Miller A Jumping Mouse (Xapm igsiynis). Cranial Measurements of Six Specimens of Zapus insignia. Number 387 1 * 1194 1195 1 4 r, 2 1 4 fi 9 Sex Q f? rT r? rT 9 Basilar length 18.5 19 19 19.8 20 19 8 Basilar length of Hensel 16.8 16.8 17.4 17.8 17.8 17.4 Zygomatic breadth 12.4 12.2 12.8 13 i:> 12.8 Mastoid breadth 10.2 10.3 10.6 1] 11 10.6 Interorbital constriction 4.8 5 5 . 5 5 5 Greatest length of nasals Incisor to molar 9.2 6 9 6 5 9.8 6 10 6 10.6 6.4 9.4 Incisor to post-palatal notch . . . Foramen magnum tq post-pala- tal notch 8.8 7 8 8.8 7 8 9 8 4 9 8 8 9 8 8 9 8 6 Upper molar series along crowns. Basioccipital to middle of pari- etal 3.7 7.f> 3.7 8.5 4 7.4 4 7.8 3.8 7.4 3.8 8 Fronto-palatal depth at middle of molar series 12 6 6 6 6.4 6 2 Greatest length of mandible 12 11.8 12 12.4 12.8 12.2 Lower molar series along crowns. Collection of E. A. and O. Bangs. Cranial Measurements of Six Specimens of Zapus hudsonius. Number 519 540 541 542 543 .547 Sex ... Q Q Q Q Q rT 1 Basilar length 18 2 18.2 17 4 17 17 17 Basilar length of Hensel Zyo'omatic breadth 16.2 11 2 16.4 11 8 15.4 11 2 15 10.8 16.2 11 15.8 11 2 Mastoid breadth 10 10 10 9 8 10 10 Interorbital constriction . ... 4 4 2 4 4.2 4.2 4 4 Greatest length of nasals Incisor to molar 8.6 6 9.2 5.4 8.6 6 8.2 5 2 8.6 6 8.6 5 2 Incisor to post-palatal notch .... Foramen magnum to post-pala- tal notch 8.4 7 4 8.2 8.2 8.4 7 4 7.8 8 8.2 8 8 8 Upper molar series along crowns. Basioccipital to middle of pari- etal b 8.2 3.2 8 3 8 3 8.2 3 7 4 3 8 Fronto-parietal depth at middle of molar series Greatest length of mandible .... Length of lower molar series alon " crowns 5.6 11.4 3.4 5.8 11.8 3 2 6 11 3 2 5.8 11.4 3 4 6 11 3 9 5.8 10.4 3.4 VOL. VIII, PP. 9-40, PLS. I-VIII JUNE 3, 1893 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON. PALEONTOLOGY OF THE CRETACEOUS FORMATIONS OF TEXAS THE INVERTEBRATE PALEONTOLOGY OF THE TRINITY DIVISION .* BY EGBERT T. HILL.f CONTENTS. Page. I. Stratigraphic Divisions and Nomenclature of the Comanchc Series 9 II. Position and Characteristics of the Trinity Division 12 III. Fossils of the Trinity Division , 15 IV. Age and Significance of the Trinity Division 17 V. Description of Species 20 VI. Plates.. 40 I.STRATIGRAPHIC DIVISIONS AND NOMENCLATURE OF THE' COMANCHE SERIES. It has heretofore been impossible to present faunal studies of the paleontology of the N various horizons of the Comanche Series, * Presented at a meeting of the Biological Society of Washington, held January 28, 1893, and published by permission of the Director of the United States Geological Survey. t The writer desires to express his indebtedness to the many friends whose advice has assisted him in the preparation of this paper : to Messrs. T. W. Stanton, W. H. Dall, Alpheus Hyatt, and F. H. Knowlton, of the United States Geological Survey, for aid in the determination of doubtful genera, and to Mr. J. L. Ridgway, artist, and Mr. C. W. Eddy, of Ware, Massachusetts, photo-engraver. 2- BIOT.. Soo. WASH., VOL. VIII, 1803. (9) 10 Hill Paleontology of tlie Trinity Division. owing to the fact that the fundamental problems concerning the sequence and relative importance of its subdivisions had not been presented until lately, although the identity of the scries, as a whole, was made known in 1886. Prior to that time most of its fossils had been described by Shumard, Roomer, and others, but it was supposed that the species all came from beds which were in some manner equivalents of the upper Cretaceous or the well known Meek and Hayden section. Since the writer ascer- tained that the Comanche Series was a distinct and lower Cre- taceous formation he has spent several years in studying the subdivisions and their extent, in ascertaining the stratigraphic position of the fossils or faunas already described, and in ar- riving at a rational system of nomenclature.* These steps were necessary before the homotaxy.of the series could be discussed. In early papers by the writer ,f pending more minute study of details, the Comanche Series was broadly divided into two divisions or convenient groups of strata, as follows : (1) An Upper or Washita Division, so named because of its prevalent occurrence in the vicinity of Old Fort \Yashita, Indian Territory, whence some of the species, which I ascertained were peculiar to this division, were originally described by Professor Jules Marcou and the brothers Shumard. (2) A Lower, or Fredericksburg Division,'! so named because many of its char- acteristic species were those described originally from the vicinity of Fredericksburg, Texas, by Dr. Ferdinand Roemer.g In 1887, while studying the Cretaceous formations of Arkansas, the writer discovered that the beds of the Fredericksburg Divis- ion, so called, consisted of two well-defined groups of strata * The Comanche Series of the Arkansas-Tex|is Region, by Robt. T. Hill : Bull. Geol. Soc. of America, vol. u, pp. 503-528. |The Topography and Geology of the Cross-Timbers and Surrounding Regions in Northern Texas : Am. Journ. Sci., vol. xxxn, April, 1887. J It has been alleged (Third Annual Report Texas Geological Survey, p. 272, and American Geologist, January, 1893), that the term Fredericks- burg Division was originated by Dr. Ferd. Roemer. Inasmuch as Dr. Roemer never recognized the existence of the Comanche Series, it is im- possible to suppose that he named its divisions, and from none of his writings can such an inference be made* \ The evolution of knowledge concerning the Cretaceous formations of Texas prior to the writer's publications is set forth in Bulletin 45 of the U. S. Geological Survey. Beds of the Comanclic Series. 11 entirely distinct from each other in paleontologic characters, and that it could be appropriately made into two divisions. For the upper of these, which is composed of the Capri na limestone' and the " Comanche Peak Group " of Shumard in part, the name Fredericksburg was retained, and to the lower the name Trinity Division was given, thus dividing the Comanche Series, more in accord with its natural grouping, into three great divis- ions, instead of two, as originally proposed. Since the publication of his Arkansas report* the writer has devoted several years to the study of the Trinity Division 'in Texas and Indian Territory, and has more fully differentiated its beds from those of the overlying Fredericksburg Division, and, furthermore, ascertained that it constitutes a paleontologic and stratigraphic division of the utmost importance in the inter- pretation of the North American Cretaceous. In the same report upon the Arkansas Cretaceous a prelim- inary description of the Trinity Division was made, separating it into a lower or arenaceous terrane, and an upper or calcareous terrane, for which, in a final paperf upon the stratigraphic sub- divisions of the Comanche Series, read before the Geological Society of America -at its Washington meeting, December, 1890, the names of the Trinity sands and Glen Rose beds were re- spectively proposed and the following general arrangement of the series given : C. The Washita or Indian Territory Division. 10. The Denison beds. 9. The Fort Worth limestone. 8. The Duck Creek chalk. 7. The Kiamitia clays or Schloenbachia beds. B. The Fredericksburg or Comanche Peak Division. 6. The Caprina and Goodland limestone. 5. The Comanche Peak chalk. 4. The Grryphaea rock and Walnut clays. 3. The Paluxy sands. A. The Trinity Division. 2. The Glen Rose, or alternating, beds. 1. The Trinity, or basal, sands. * The Neozoic Geology of Southwestern Arkansas, vol. u of the Annual Report, Geological Survey of Arkansas, 1888. t Loc. cit. 12 Hill Paleo-htblogy'ofthe Trinity />/V/.s/o//.. IL POSITION AND CHARACTERISTICS OF THE TRIMTY DIVISION. Fuller details concerning the extent and occurrence of the Trinity Division have been largely set forth during the past year in a publication by the writer,* and much of its stratigraphic details and some of the final maps as made by him and his as- sistants, Messrs. J. A. Taff, J. S. Stone, W. T. Davidson, and N. F. Drake, have been printed still later.f The details of the formation can be appreciated by referring to the published descriptions of three typical sections, made at widely separated intervals, showing the increasing thickness of the beds to the southward. The first of these, made by the writer while employed upon the work of the Arkansas State survey, is described in the report upon the Neozoic geology of southwestern Arkansas. J The second represents the beds as they occur in the vicinity of the Paluxy and Brazos rivers, in north central Texas, and was made by the writer and his assistants, Messrs. J. S. Stone and W. T. Davidson, and published in the report " On the Occur- rence of Underground Water," and later in the Third Annual Report of "the Geological Survey of Texas. || In this region the beds of the Trinity Division attain an established thickness of about 475 feet. The third and southernmost section was made along the banks of the Colorado river between Austin and the Paleozoic contact in Burnet county by the writer and his assistants, Messrs. J. A. Taff and N. F. Drake. The rocks of the Trinity Division attain unusual development in this region and are very satisfactorily exposed. A profile ^j illustrating this section is published in the *"On the Occurrence of Artesian and Other Underground Waters in Texas, New Mexico, and Indian Territory West of the Ninety-seventh Meridian" (pp. 41-166 of " Final Reports of the Artesian and Underflow Investigation, etc., to the Secretary of Agriculture." 52d Congress, 1st Session, Ex. Doc. 41, part 3, Washing-ton, D. C., May, 1S ( J2). t Report on the Cretaceous Area North of the Colorado River, by J. A. Tuff. Austin, Texas, September, 1892. tOp. cit, pp. 116-126. Op. cit., pp. 111-112. || Op. cit, pp. 307,310,311. ^[Loc. cit, pp. 90-91. The. Trinity Division. 13 report upon the " Occurrence of Artesian Water," and much of the detail of the section given in the Third Annual Report of the Texas Geological Survey.* The thickness of the rocks of the Trinity Division in this region is about 500 feet. The beds have not been systematically studied in their extent south of the Colo- rado river. From study of these sections it is concluded that the beds, as a whole, indicate a progressive and continuous series of sedi- ments, representing subsidence from land through littoral to off- shore conditions, followed by renewed shallowing at its close. It consists of sands and conglomerates at its base, and grades upward into magnesian and chalky limestones. No sharp lines of demarcation can be drawn between the sands and limestones, so imperceptibly do they merge into each other. The Basement Beds or Trinity Sands proper. These consist mostly of unconsolidated fine conglomerate and sands of the nature locally known as pack sands, and contain, besides logs of silicified wood, occasional masses of firm, lustrous lignite of depressed oval cross-section, like those found in the Potomac formation near Muirkirk, Maryland. Large bones of vertebrates have also occasionally been found, notably near Millsap, Texas, Travis Peak post-office, and at Gypsum Bluffs, Arkansas, which are supposed to be the remains of Dinosaurs. It was owing to the occurrence of these bones that in an early paper these sands were at first termed the Dinosaur sands by the writer.f The Glen Rose Beds. Indurated layers of impure calcareous and yellow material succeed the sands, and become more calca- reous and magnesian toward the top of the sections, but without any defined breaks in the sedimentation. In the medial and upper portions of the sections the magnesian and limestone strata assume great thickness and purity, .and are separated by alternations of laminated, calcareous, and magnesian clays, as beautifully shown in the bluffs of Mount Bonnel, on the Colo- rado river, northwest of Austin. J *Loc. cit., pp. 265-300; also First Annual Report of Texas Geological Survey for 1889, 1890, p. Ixxxv. f Am. Journ. Sci., vol. xxxin, April, 1887, p. 298. \ Photographs of the scenery and structure of the Cretaceous forma- tions of Texas, made by the writer for the Texas State Geological Survey, can be procured from the Committee on Photographs of the Geological Society of America. 14 Hill Paleontology of the Trinity Division. Molluscan and other invertebrate remains appear coincident with the calcareous beds, accompanied in some instances by plant and vertebrate remains, as at the plant beds three miles west of Glen Rose, Somervell county, Texas. Aggregations of Species in Great Beds. In various parts of the Glen Rose beds there are strata composed of shells of one pre- dominant species, while in other cases there is an agglutination of shell fragments of many species in masses similar to the recent formation on the coast of Florida known as Coquina. Coquina Beds. These usually appear at the base of the Glen Rose beds or at the first appearance of marine mollusks in the series. In Arkansas, owing to greater alteration through cal- cification, they consist of much more indurated limestone mate- rial than in Texas. The massive beds are composed almost ex- clusively of small shells of many species, and usually have a dark-yellow color upon weathering. They outcrop at many places along the old military road between Antoine and Ultima Thiile. Shell beds are especially well developed near Travis Peak post-office, near the Colorado river, where the Coquina beds are pure white in color and the shell fragments more sili- ceous and comminuted than in Arkansas. The Ouster Agglomerate. Near the base of the Travis Peak sec- tion is a stratum some four feet in thickness, composed exclu- sively of a fossil Ostrea, so poorly preserved that the specific nature cannot be ascertained; but which resembles 0. franUlnl Coquand. A similar bed of Ostrea franklini occurs in the west bluff of the Little Missouri, three miles west of Murfreesboro, Arkansas. The Vicarya Beds. At Post Mountain, west of the town of Burriet, there is the remnant of a vast bed of agglomerate, com- posed entirely of the shells herein described as Vicarya lujani de Verneuil, cemented by a hydrocarbon matrix, probably gra- hamite. This bed is some ten feet in thickness, and is evidently near the base of the Glen Rose beds. The Orbitidites Chalk. Near the base of the Bluffs of the Col- orado, about the middle of the Glen Rose beds (Upper subdivis- ion) is a stratum of ten feet or more in thickness, composed entirely of a massive white chalk, studded with the minute shells of the foraminifera Patellina (Orbituliteti) texana Roemer. This chalk extends southward into Hays. Comal, and adjacent coun- ties. Fossils of the Trinity Division. 15 The Requienia (''Caprotina") Limestone. At Granbury and in southwestern Parker comity there is a vast agglomerate com- posed entirely of the shells and casts of Requienia texana Roomer, to which Dr. B. F. Shumard gave the name "Caprotina lime- stone." This occurs well up in the Glen Rose beds, about one hundred feet from their top. There is a similar bed in the ravine near the east foot of Mount Bonnel, west of Austin. As shown by the writer,* there is another horizon of Caprotina limestone higher in the Comanche Series. Th& Nerinfea Flags. The summit of the Glen Rose beds, as seen in Mount Bonnel, consists of alternations of dimension layers of firm crystalline limestone and pseudo-oolitic marls. Some of these dimension layers in the peak of Mount Bonnel are composed almost exclusively of calcined forms of Nerinteas. III. FOSSILS OF THE TRINITY UIVISION. The bods contain many plant, vertebrate, and invertebrate remains, which occur either separately or in association, as at one locality in the bed and banks of Paluxy creek, three miles west of Glen Rose, Texas. The more sandy basement beds, or Trinity sands, are, as a rule, deficient in organic remains, with the exception of silicified wood and lignite, although occasional vertebrates and invertebrates are found. The lower beds of the Glen Rose subdivision abound in inverte- brate remains, most of which, however, are but badly preserved casts. The Glen Rose beds also contain occasional plants and vertebrates, especially in their lower portion, where they grade into the underlying sands. The upper beds of the Glen Rose Division are less abundant in fossil remains, and these are very poorly preserved as casts. The plants of the basement Glen Rose beds have been collected by Professor Lester F. Ward, of the United States Geological Survey, and studied by Professor W. M. Fontaine, of the Uni- versity of Virginia. They are now in course of publication in the Proceedings of the United States National Museum. f The small but interesting collections of vertebrate remains, with the exception of a lepidotoid fish in the hands of Professor Cope, * First Annual Report of the Geological Survey of Texas, 1889, Aus- tin, 1890, p. 133. fProc. U. S. Nat. Mus., vol. xv, pp. , pis. xxxv-Lxrn. 16 Hill Paleontology of the Trinity Division. have not been studied. In this paper will be presented a pre- liminary study of its invertebrate fauna. The following is a list of the fossils collected or observed by the writer from all the beds of the Trinity Division in Arkansas and Texas : Foraminifera : Patetlina texana (Roemer). Echinodermata : Epiasler (?) sp. inclet. Vermes: Serpula paluxiensis sp. nov. Molluscoidea : Genus indeterminate. Mollusca: Anomia texana sp. nov. Ostrea franklini Coquand. Ostrea franklini ragsdalei var. nov Pecten stantoni sp. nov. Modiola branneri sp. nov. Leda harveyi sp. nov. Cucullsea graliota Hill. Cucullxa comanchensis sp. nov. Cucull&a terminalis Conrad. Barbatia parva-missouriensis Hill. Trigonia stolleyi sp. nov. Trigonia crenulata Roemer. Chione (?) decepta s'p. nov. Eriphyla pikensis Hill. Requienia texana (?) (Roemer). Monopleura marcida White. M. pinguiscida White. Corbicida arkansaensis Hill. Cardiam (?) sevierense Hill. Protocardia sp. indet. Pholadomya knowltoni sp. nov. Pholadomya ierchi sp. nov. Pteuromya (?) henselli sp. nov. Isocardia (?) medialis Conrad. Natica (?) texana Conrad. Age of the Trinity Division. . 17 Viviparus natica (?) cossatotensis Hill. Cylindrites (?) sp. indet. Buccinopsis (?) parryi Conrad. Tylostoma pcdernalis (Rocmor). Vicarya branneri sp. nov. Nerinsea austinensis Roemer. Neritina sp. indet. Neumayria walcotti Hill. Acanthoceras (?) Justin fp sp. nov. Arthropoda : Cypridca texana sp. nov. Vertcbrata : Lcpidotus. Crocodila. Dinosauriia. CJidonia. Plantse : Many species, now in course of publication by Professors Ward and Fontaine, together with undetermined species resembling Araucarites, figured on plate 1. IV. AGE AND SIGNIFICANCE OF THE TRINITY DIVISION. With the exception of the genera Requienia and Monopleura, the above list contains none of the hitherto familiar types of fos- sils found in the overlying Eredericksburg and Washita Divisions, such as the characteristic Echinodermata, Radiolites (alleged Hip- purites), Gryphxas of the Pitcheri group, Exogyras, or Schloenbach- iale Ammonitidsz, but possesses a molluscan fauna peculiarly its own, so far as America is concerned, which, accompanied by a well-preserved flora and vertebrate fauna, affords the first satis- factory and complete data for an age classification of the sub- divisions of the North American Lower Cretaceous formation, and will enable us to more thoroughly interpret the succeeding divisions. This association of vertebrates, plants, and marine mollusca is a most important fact in the correct determination of the age of these beds, and they all apparently agree in con- clusions. Concerning the interpretation of the foregoing fossils, the fol- lowing facts may be stated : The plants, as determined by Pro- 3-BioL. Soc. WASH., VOL. VIII, 18915 18 Hill Paleontology of the Trinity Division, fessors Ward and Fontaine, originally in the Potomac region,* were long since referred by them to the Wealden', before they were known to occur in the magnificent Texas stratigraphic series. The vertebrates have never been systematically studied nor collected, but the genera found all occur in the Lower Creta- ceous of Europe. Of the invertebrates the ostracoid crustacean Cypridea of the Glen Rose lower beds is, in Europe as in Texas, a prominent feature- of the Wealden (Lower Neocomian) beds. The foraminifer Patellina (Qrbitulites) texana Roemer is indis- tinguishable from the Orbitulites, which characterizes the Upper Neocomian of France, and occurs there under lithologic condi- tions similar to those in Texas. Of the Pelecypod mollusca proper the Anomia is indistinguish- able from the Anomias of other ages. The only Ostrea (0. franldini Coquand) seem identical with the figures of a form which has been described under many spe- cific names from the Upper Jurassic, and Lower Cretaceous of Europe. Pecten stantoni belongs to a group of the PectinidiK, which has great specific development in the Neocomian of France. Spain, and Portugal. Modiola, Leda, Cuculltea, Protocardia, Corbicula, Pholadomya, and the doubtful form called Isocardia have a wide range in the geologic column, but the forms found in the Trinity Division have a general varietal resemblance to those of the Neocomian. The only well-defined species of Trigonia (7 1 . stolleyi sp. nov.) belongs to the scab rate forms peculiar to the Cretaceous and later epochs. This is an important fact against the possible Jurassic age of the beds. The aberrant genera, Requienia and Monopleura, which abound in the Glen Rose beds,' are both characteristic Lower Cretaceous genera, occurring abundantly in Europe in the Neocomian beds and not ranging higher than the Cenomanian. . In describing Requienia (Caprotina} texana, Roemer asserted that it was hardly distinguishable from the characteristic Caprotina lonsdallii of the Neocomian of France. Of the Gastropoda, the genus Vycaria is represented by V. branneri resembling a peculiar species of the European Neo- * See various papers by Professors Lester F. Ward and W. M. Fontaine on the Potomac flora of the North Atlantic coast. Age of the Trinity Division. 19 coinian, V. lnjani De Verneuil and which in every variation is identical with the figures of the European species. Natica (Tylostoma) pedernalis Roemer is characteristic of the Tylostomas of the Neocomian of France, Spain, and Portugal. Of the Nerinasas in the Texas beds all have the archaic form of the Jurassic and lower Cretaceous (Neocomian} Nerinreas. Only one echinoid is found in the Trinity Division, Epiaster (?), but this is of the older Cretaceous aspect of the European forms. Of the Ammonitidae, which in Europe are most relied upon for the classification of subdivisions of the Cretaceous, it may be said that the Comanche Series below the Washita Division is very deficient in these, only four species being known in America. Of two of these only three individual specimens have been found, while the European Neocomian abounds in many species and genera. Of the two genera with one species each found in the Trinity Division, it may be said that one of them, Neumayridj belongs to a genus which occurs in the Pur- beckian, or uppermost Jurassic, and Wealden of Europe, and hence may be accepted as strong evidence that these beds are not of late Cretaceous age. The other species, Acanthocems (?) justinae, is too poorly preserved to be of criterional value. While the writer has throughout placed the Trinity Division in the Cretaceous, he tried to defer final discussion of their age until opportunity should arrive for careful study of these fossils. Owing to constant labors in the field upon the more important* stratigraphic problems, this opportunity did not arrive until now. At the time the Arkansas report* was written it was held that the Trinity beds might prove to be Jurassic, but the careful revision here presented tends to remove this doubt and enables us to assert their Cretaceous age with more assurance. What- ever doubt may have been inferred from any expressions in previous publications,! it may now be stated positively : * Op. cit. f Through two unfortunate lapses in the typography of his former papers the attempt has been made to show that the writer did not hold the Cretaceous age of the Trinity Division, notwithstanding his repeated publications to the contrary. One of these is caused by the typographic error on page 84 of the report " On the Occurrence of Underground Waters," etc., Washington, May, 1892, where the clause i% which are as- sumed to be the base of the true Cretaceous " is made to modify the words " Walnut clays," instead of " these beds," i. e., the Trinity. The other lapse was of a somewhat similar nature in the previous publication on The Comanche Series of the Texas-Arkansas Region. 20 Hill Paleontology of the Trinity Division. 1. That there is not a single invertebrate species in the Trinity Division of exclusive Jurassic age, which would justify placing the beds in that period. 2. The genera all occur in the Cretaceous formations of the rest of the world, and many of them, such as Requienia and Monopleura, occur only in the Cretaceous. Hence the beds are Cretaceous. 3. The beds of the Trinity Division are of lowest Cretaceous age, Neocomian, because the genera all occur in the Neocomiaii or lowest Cretaceous of other countries, and because they con- tain none of the characteristic upper or middle Cretaceous forms. Finally it may be stated that from the above comparison of the life of the Trinity Division with the Cretaceous life of Europe it is evident that it shows not only a resemblance, but a re- markable homotaxial similarity with the Lower or Neocomian of that country, the lowest faunas resembling the Wealden or Lower Neocomian, and the Upper Glen Rose beds the Middle and Upper Neocomian, especially as developed in the region of the Jura and in Spain and Portugal. V. DESCRIPTIONS OF SPECIES. FOR AMINIFER 7E. Patellina texana (Koemer). Plate I, Figs. 2 (copied after Roemer), 2a, 26, 2c, 2d. Orbit-unites texana Roemer. Die Kreidebildungen von Texas, p. 86, plate x, figs. 7a, b, c, d. " Shell minute, attaining one-eighth of an inch in diameter ; orbiculate, shield-shape, convex above, obtusely conical ; central eminence umboniform, ornamented with close, fine, concentric stria3, otherwise smooth ; lower part fiat, slightly concave, with irregular, radiating, granular ruga), as if perforated by worms. "- Roemer. Roemer said : " This species certainly belongs to that group of Lamarck's genus Orbitiditcs, which D'Orbigny separated as a distinct genus Orbitolina." Careful microscopic study of the interior structure by the writer shows the granular structure illustrated in figs. 2a, 26, and that it belongs to the genus Patellina of Williamson. Descriptions of Species. 21 This important foraminifer was first found by Roemer on the upper branch of the Pedernalis, associated with Tylostoma pcder- nalis. I have found its true position to be in the medial por- tion of the Glen Rose beds, where, as on the Colorado near the mouth of Bull creek, it occurs in a massive chalk some ten feet thick. The "Orbitolina " beds are among the most characteristic and distinguishing features of the upper Neocomian of Europe, in the region of Jura, and in Dauphine, Portugal, and Spain. Their occurrence in a similar stratigraphies position in the Texas region is additional evidence of the Neocomian age of the Glen Rose beds. This Texas form has also been reported by Karsten from the Neocomian beds of Venezuela. ECHINODERMATA. Epiaster (?) sp. indet. Only one echinoid has come under the writer's observation from the Glen Rose beds. It occurs about midway in the sec- tion, near Bull creek, Travis county, and in Somervell county, Texas. This has been submitted to Professor W. B. Clark, of Johns Hopkins University, who writes as follows concerning it: " I am inclined to think it not only a new species, but a new genus." The specimens are left in Professor Clark's hands for future determination. VERMES. Serpula paluxiensis sp. nov. Plate I, Figs. 4, 4a, 46. Cylindrical tubicolate, marked by concentric lines of growth; occurs in colonies, radiating out from a central nucleus. Indi- viduals several inches long, but it is impossible to trace com- plete length of specimens, owing to mode of growth. Adult specimens average one-eighth inch in diameter. This is one of the most abundant forms in the basement hori- zon of the Glen Rose beds, and occurs attached to lignite, shells of Ostreidae, and casts of other mollusks, or'iii immense colonies or spherical masses, some of which in Paluxy creek, averaging three feet in diameter, are composed entirely of this species. 22 Hill Paleontology of the Trinity The Serpulas have such wide geologic range and -so few specific characters that they are of little value in geologic diagnosis. Occurs at gypsum bluffs of the Little Missouri, and in great abundance at. the plant bed near Glen Hose, and also throughout the extent of the lower fossiliferous Glen Uose beds in Texas. MOLLUSCOIDEA. Genus indeterminate. Microscopic oval cells about one millimeter in length, growing in colonies attached to shell ofSerpula palttxiensrs and other forms ; cells not overlapping, but in close contact with each other, form- ing a single layer of delicate net-work. The cells of this species have not the pyriform shape or im- bricated arrangement of Membranipora or the vibracular cells of LunuliteSj and hence are assigned to no generic position at pres- ent. This form is the only one belonging to this order yet found in these beds. It occurs attached to other shells in the beds of the plant locality on the Paluxy near Glen Rose, Texas, at the base of the Glen Rose beds. MOLLUSCA. Anomia texana sp. nov. Plate I, Fig. 5. Anomia sp. indet. Hill. Arkansas Geological Survey, Annual Report 1888, vol. n, p. 135. Thin, discoidal, indistinguishable specifically from many spe- cies of this genus ; right or lower valve attached, concentrically laminated; left upper valve arched and very irregular; seldom exceeds one-half inch in greatest diameter. This species abounds in the earliest fossiliferous horizon of the Trinity Division, such as the beds in Paluxy creek, west of Glen Rose, and at the gypsum bluffs of the Little Missouri, in Pike county, Arkansas. It also occurs in most of the localities throughout the extent of- the Glen Rose beds. Descriptions oj Species. 23 Ostrea fraiiklini Coquand. Monographic du Genre Ostrea, p. 58, plate xxiii, figs. 8-10. Hill, Arkansas Geological Survey, Annual Report 1888, vol. n, plate v, figs. l-18a; plate vi, figs. 19-25 ; plate vii, figs. 28-30. The general aspects and variation of this characteristic oyster of the Trinity Division have been fully described and figured in my Arkansas report. Professor Marcou in a review of this work* has divided the form into many species, but the writer, from his extensive study of the occurrence of the specimens in situ, still believes in the unity of the species, although in Europe it has doubtless been the custom of earlier paleontologists to make many species out of variations. ChofFatf figures and describes from Portugal 0. barrosei, a form which resembles a variety of 0. franklini found at Glen Rose. This species occurs in great abundance throughout the Trinity Division, especially at the plant beds near Glen Rose, at the base of the Glen Rose beds in the Colorado section, and in a similar horizon throughout the Glen Rose beds in Arkansas and in Texas. It is interesting to note that none of the true Grypk&ds or Ewgyras have yet been found in the Trinity Division. Ostrea franklini ragsdalei var. nov. Plate I, Fig. 6. Shell acuminate, oblong, marked by numerous, regular longi- tudinal coske ; beak of large valve prolonged, costate, sub-cylin- drical. Several incomplete specimens of the larger valve of this spe- cies were procured from the fauna at the plant bed near Glen Rose. The outline is somewhat similar to that of 0. franklini Coquand, but the larger valve is much more round, the point more prolonged and characterized by the strong costac which do not appear upon the adult specimens of the 0. franklini else- where found. * American Geologist, vol. iv, December, 1889, pp. 359, 360. f Recueil de Iconographies Stratigraphiques Sur le Systeme Cretacique du Portugal, par Paul Choffat. Lisbon, 18S5, p. 37, plate iii, figs. 7, 8, 9, 10, 11, 12. 24 Hill Paleontology of tJie Trinity Division. This variety is named for Mr. G. H. Ragsdale, the naturalist, of Gainesville, Texas. Thus far this variety has only been found at Glen Rose. Pecten stantoni sp. nov. Plate II, Fig. 3, 3a. Shell small, one and one-half inches in length, strongly eared, and ears rugose, the right one (not shown in figure) being marked by a deep fold. The surface of the larger valve is marked by strong, flattened, double ribs, each with a sinus its entire length, alternating with small single ribs, and by minute cross-lines. This marking distinguishes it from the FoZa-like forms of the upper half of the Comanche Series. The smaller valve is not known. This beautiful species occurs in the molluscan fauna at the plant bed on Paluxy creek, near Glen Rose ; only three speci- mens have been found. Two of the specimens ,were very perfect, but were unfortunately lost in the removal of my collections from Cornell University. This species is named for Mr. T. W. Stanton of the U. S. Geo- logical Survey. Modiola branneri sp. nov. Plate V, Figs. 8, 9, 10. Modiola sp. indet. Hill. Arkansas Geological Survey, Annual Report 1888, vol. n, p. 133, plate ii, figs. 18, 19. Shell small, elongate, from one-half to one and three-quarters of an inch in length, elongated sub-triangulate, greatly thickened at umbonal region ; umbones pronounced and rapidly narrowing to a rounded point; anterior portion somewhat flattened ; pos- terior portion attenuated, thin, and strongly curving in outline ; surface smooth, lustrous, marked by fine lines of concentric growth and faint radiating striae. This well-preserved little Modiola occurs sparingly in the lower Glen Rose beds at the gypsum bluffs of the Little Missouri, Arkansas, and in the plant bed near Glen Rose. It was orig- inally figured, but not named, in my Arkansas report. Descriptions of Species. 25 Leda (?) harveyi sp. nov. Plate I, Figs. 7, 8. Shell minute, one-sixteenth to one-eighth of an inch in length, elongate, smooth, concentrically striate; beaks situated at ante- rior third, blunt, thick, and recurving ; posterior cardinal margin elongate, straight, or slightly concave ; pallia! margin straight or slightly sinuous ; anterior cardinal margin short, straight ; ante- rior margin very slightly rounded at base ; posterior margin angular, short, and straight; anterior half of shell thick, sub- globular; posterior half elongated, rapidly thinning posteriorly, and sometimes marked by a strong angular ridge extending from beak to union of pallia! and posterior margin ; cardinal area not exposed so as to show hinge mechanism. This minute shell occurs in great masses, resembling small black specks in a calcareous cement, which under the magnify- ing glass reveals the outline shown in the figures ; the hinge mechanism not seen. Found thus far only at the plant beds of the Paluxy, near Glen Rose. Named in honor of Mr. J. W. Harvey, who first collected from this locality, but died before the collections could be published.* Cucullaea gratiota Hill. Area gratiota Hill. Arkansas 'Geological Survey, Annual Re- port 1888, vol. IT, p. 133, plate *4^iigs. 2, 2a. Described and figured in my Arkansas Report as Area gratiota. Occurs also in plant beds near Glen Rose, Texas, and in bluffs of Colorado. Cucullaea comanchensis sp. nov. Plate III, Figs. 1, 2. Cordate, globose, thicker and higher than long, subquadrate in lateral aspect ; posterior margin strongly truncate, pallial margin gently rounded; anterior margin short and truncate; exterior of cardinal area elongate and broad ; umbones small, high, and incurved, but not touching; shell thick, rough, and marked by strong, irregular, rugose lines ; hinge mechanism not visible. * See American Geologist, October, 1802. 4 BIOI,. Soc. WASH., VOT,. VIII, 1893. 26 Hill Paleontology of the Trinity Division. This is one of the most characteristic species of the very base of the Glen Rose beds in Parker, Hood, and Comanche counties, and is the form which the writer once inferred to be analagous to Ambonicardia cookii Whitfield,* from the Raritan clays of New Jersey, which the fragments then found strongly resem- bled in outline and marking. The discovery of the specimen here figured, however, for the first time affords evidence for their satisfactory generic identification. The specimen figured was found about three miles east of Millsap. Texas, at the contact of the Trinity sands and the cal- careous Glen Rose" beds. I have also found the species near Springtown, Parker county, and at Comanche, in the same hori- zon. Cucullaea terminalis Conrad. . (See Report United States and Mexican Boundary Survey 1 Washington, 1857, vol. i, p. 148, plate iv, figs. 2a and />.) Casts of this species are frequently met with in the .Glen Rose beds. The shell has not yet been found. Barbatia parva-missouriensis Hill. This species, from Pike county, Arkansas, was described and figured in my Arkansas Report, p. 133, plate iv, figs. 4o, 4/>, 5, and probably fig. 22, of plate ii, of same report. Trigonia stolleyi sp. nov. Plate III, Figs. 3 and 5. Semi-lunate in general outline, beaks well forward and strongl} 7 recurved ; anterior and pallial margin a strong continuous curve ; posterior portion elongated with truncated posterior margin ; cardinal area compressed. Surface marked by flexuous, nodulcd costa), about twenty -two in number, narrow and high, separated by broad intercostal areas as in T. alfeformis Link. ; depressed cardinal area bordered on its outer side by a long narrow groove and marked by cross-ribs, ^flexing anteriorly. This Trigonia differs from T. emoryi Conrad, of the Washita Division, in its general outline and entirely distinct surface mark- *See Report of Arkansas State Geological Survey for 1888, vol. n, p. 126. Descriptions of Species. 27 ing. (See U. S. and Mexican Boundary Report, vol. i, p. 148, plate iii. figs. 2, 6, c.) Occurs abundantly in the plant beds of the Paluxy. The species is named for Professor G. W. Stolley, of Austin, Texas, whose work as a collector is esteemed in both Europe and America. Trigonia cienulata Roerner. Plate III, Fig. 4. Roomer. Kreidebildun .>;en von Texas, p. 51, plate vii, fig. 6. This species, described by Roemer from imperfect casts, may be the same as T. stolieyi. The specimen here figured was col- lected from the bluffs of the Colorado near Bull creek, Travis county. Chione (?) decepta sp. nov. Plate I, Figs. 9 and 10. Shell sub-triangulate or elliptical in outline, compressed in cross-section from base to um bones ; umbones forward of center, small, and touching each other ; lunular area small, cordate oval ; pallial margin widely rounding ; anterior cardinal margin con- cave, shorter than posterior ; posterior cardinal margin long, slightly con vex, asymmetrical ; cardinal mechanism not shown; surface marked by long concentric and very faint irregular lines, anterior muscular scar elongate, rounded ; posterior muscular impression flat, depressed at extreme posterior end. The casts of this form are abundant throughout the Glen Rose beds, but its generic position is uncertain. It may possibly be- long to the Carditidse. In the hills north of Lampasas, near the top of the formation, the specimens occur in great abundance with the shell preserved in calcite, showing the exterior structure. Conrad's species of Astarte tcxana* described from a cast from an unknown locality, has a superficial resemblance to this form, according to his description and figures, but it is more triangular and otherwise different, as seen by comparison with his type in the National Museum. * U. S. and Mexican Boundary Report, vol. i, p. 152, plate v, fig. 9. 28 Hill Paleontology of the Trinity Eriphyla pikensis Hill. Plate IV, Figs. 4, 5, 6. (hrliiciila pILrn.*!* Hill. Arkansas Geological Survey, Annual Import ISSK, vol. n, p. i:n, plate ii, figs, Hi, l&J, M, 15, 1C,. 17. This form was originally figured by the writer under the ge- neric name of Corbicula, and is very closely allied to the so-called Cyrena astarteformis Koch and Dunker, from the Wealden of Germany. Professor Marcou has referred it to the Astartidse,, but it is undoubtedly a species of the genus Eriphyla of the AstartidaB distinguished by the lateral teeth. It occurs in great abundance in Pike county, Arkansas, and sparingly at-the plant bed of the Paluxy, near Glen Rose, Texas. Requienia texana (?) (Roemer). Caprotina texana Roemer. Kreidebildungen von Texas, p. 80, plate v, figs. 2a, 26. A Requienia, provisionally referred to R. texana Roemer, is one of the prominent species of the Glen Rose beds and occurs in massive agglomerate some twenty feet in thickness at Glen Rose, Thorp Springs, Granbury, and in southern Parker county. In Roemer's description he asserts that it is indistinguishable from Requienia lonsdalli D'Orb., of the French Neocomian, except by its thicker shell. It is desirable to closely compare this form with R. patagiata White.* Dr. Roemer says that the latter spe- cies is entirely distinct, the larger valve of R. texana not being so elevated as in R. patagiata. In the abundant material col- lected by the Avriter this distinction does not always hold good, yet there is a general difference in appearance, especially in the larger size and more rounded character of the valves of the Glen Rose forms, which may make it a distinct species from either of these. The type forms of the R. texana Roemer and the R. patagiata White occur in the horizon which we at present accept as the Caprina limestone, while the Requienias of the agglomerate at Granbury and at the base of the Colorado section occur several hundred feet below them, and may prove a distinct species. Requienia is the lowest occurring genus of the aberrant Cham- '* Requienia patagiata White. U. S. Geological Survey, Bulletin No. 4, p. 6, plate v, figs. 1-8. Descriptions of Species. 29 icUc, with the exception of Diceras, a Jurassic form, and is abun- dant in the Neocomian formation of Spain, France, and Portugal. Monopleura marcida and M. pinguiscula White. Bulletin U. S. Geological Survey, No. 4, p. 8, plate 5, figs. 1-8. Casts and moulds of Monopleura occur in great abundance in the Colorado section, especially at the base of Mount Bonnel, near the mouth of Bull creek. It is impossible to make a cor- rect diagnosis of these, owing to the fact that the smaller valves, so abundant in the Caprina limestone, have not been found in the Glen Rose beds, but there is a general resemblance of the larger valves to the two species named. The genus Monopleura, according to Zittel, has wide distribu- tion in the Lower Cretaceous, occurring in the Neocomian of Provence and the Jura mountains, and seldom ranging higher. In America the genus culminates in the Caprina limestone and is not known later. Corbicula arkansaensis Hill. Arkansas Geological Survey, Annual Report 1888, vol. u, p. 133, plate ii, fig. 20; plate iv, figs. 3, 3, 6. This species was originally figured and described in my Ar- kansas report. It occurs in Pike county, Arkansas, near Mur- freesboro, and at the gypsum bluffs of the Little Missouri. It is less abundant in the Texas beds. Cardium (?) sevierense Hill. [7 Arkansas Geological Survey, Annual Report 1888, vol. n, p. 134, plate ii, figs. 21, 21d. This form has only been found in Arkansas, at the locality from which it was described in my Arkansas report. Protocardia sp. indet. Small ousts, three-quarters of an inch in length, globose; sur- face marking, very fine lines. The Comanche species of the genus Protocardia need careful revision, and this form may be found to belong to some of the numerous species already described. This species is distin- guished only by a smaller size than that of all the other forms described. It occurs sparsely in the beds near Glen Rose. 30 Hill Palvo'idoloyy of tltc Trinity Division. Pholadomya knowltoni sp. nov. Plate II, Figs. 1, 2. / Choffat (Materiaux p6ur L'Etude Stratigraphique et Paleon- tologique de la Province D'Angola, Geneva, 1888 r p. 84, plate v, figs. 1-3) describes under the name of P. pleuromyaeformia a form indistinguishable from this species. His description, as follows, corresponds fully with our species : " Equi valve, inequilateral ; swollen below the beaks ; anterior bordeif rounded and completely closed ; posterior border com- pressed at the extremity, which is slightly turned upward, trun- cated and slightly gaping ; beaks small, elevated, strongly innexed and in contact with each other ; cardinal portion of anterior border sloping and its continuation strongly rounded; the pos- terior cardinal border straight, slightly elevated to its extremity ; anterior face blunt, behind which a slight, faint groove extends from the beak to the pallial border. The surface of the shell is marked by irregular longitudinal plications." Choffat. This shell can in no way be distinguished from the excellent figures and descriptions given by Choffat of Pholadomya plcuro- mydeformis, from Dombey, on the west coast of Africa, where a fauna closely allied to the Comanche series occurs, but of course their identity cannot be positively established without compari- son of specimens. The faint grooves from beak to pallial border are not brought out well in our figures. The form first appears in America in the medial portion of the Glen Rose beds of the Colorado river section, near the mouth of Bull creek, and again appears in the supposed Caprina lime- stone at Austin, in the Fredericksburg Division. Pholadomya leichi sp. nov. Plate IV, Fig. 3. Outline subpyramidal in lateral aspect ; length, three and one- half inches ; height, two and one-half inches ; greatest thickness, two inches ; beak situated at anterior third, of medium propor- tions ; anterior margin semicircular in outline from beak to pal- lial margin, into which it merges by a continuous curve ; pallial margin a continuous curve with the anterior margin, and rapidly increasing in curvature posteriorward, terminating obtusely with the truncated posterior margin ; posterior margin sharply trun- cate, about one inch in length ; anterior umbonal margin very short, marked by a small depression immediately below the Descriptions of Species. 31 umbone ; posterior umbonal margin elongate, sloping posteriorly at an angle of about 30 degrees from tbe umbone ; this margin is straight when viewed laterally, but bends strongly outward toward the gaping posterior margin, with which it unites by a gentle curve; surface marked by strongly rounded ribs and grooves, subconcentric, sinuous and uniting with each other at a common groove which is parallel with the anterior margin. Only a single poorly preserved cast of a right valve of this species has been found, the anterior portion of which is not visible. It was collected in the heavy conglomerate which marks the base of the Comanche Series at its contact with the Carboniferous formation, on Sycamore creek, Burnet county, Texas, near the crossing of the Burnet and Travis Peak roads. The species is named for Dr. Otto Lerch, of the. Louisiana State Geological Corps, who, as a Texan, has made valuable contri- butions to the knowledge of his State. Pleuromya (?) henselli sp. nov. Plate IV, Figs. 1, 2. Shell elongate-elliptical ; length, three and one-quarter inohes ; height, two inches; umbones anteriorly subcentral, round, and not prominent; anterior margin strongly rounded, continuing by curvature into the pallia! margin ; pallial margin elongate, slightly curved ; posterior margin sub-truncate, slightly gaping ; anterior umbonal margin sinuous, slightly gaping ; posterior umbonal margin about one-third longer than anterior ; interior and hinge mechanism not seen. The generic position of this species is very doubtful ; but, in- asmuch as it is one of the most characteristic casts of the Glen Rose beds in the Colorado River section, it is important that it should be here figured. The species has a resemblance to Thra- cia mijcisformis White, but differs in some details. Isocardia (?) medialis (Conrad). Plate II, Figs. 4/5; plate III, Fig. 6. Cardinm mediate Conrad, U. S. and Mexican Boundary Report, vol. i, p. 149, plate iv, figs. 4a, b. Conrad described this form as follows : " Cordate equilateral, ventricose,; base profoundly and nearly regularly rounded ; beaks prominent ; posterior margin truncated , direct." 32 Hill Paleonlologij of tJtc Trinity Division. To this I would add : Variable in shape; shell thick, concentrically striate ; beaks inturned as in Isocardia; anterior muscular impression very prominent, angular on anterior side ; posterior muscular impres- sion faint, large in area, almost indistinguishable; posterior margin truncate in normal specimens. Surface marked by Strong concentric rugose lines. The generic position of this form is not satisfactory, because the hinge structure has not been found. It is clearly not a Cyprina, although sometimes similar forms are referred to that genus by paleontologists in Europe; neither is it a (.hrdiinit as described by Conrad. This is one of the most numerous, conspicuous, and character- istic species of the Glen Rose beds; it occurs asshelless casts by the thousands throughout the vertical and geographic extent in Texas and Arkansas, beginning in the lowest fossiliferous hori- zon and extending to the top. A few specimens were found about ten miles west of Glen Rose, upon which fragments of the thick calcined shell structure were preserved, showing it to have a concentrically striated sur- face and the anterior cardinal margin and beaks to be as in Iso- cardia, as shown on plate vii, fig. 1. The form varies greatly in shape, owing to the compression and distortion it has undergone in the strata. One of the ex- treme variations from the normal is illustrated on plate ii, fig. 5. Natica (?) texana Conrad. U. S. and Mexican Boundary Survey, p. 157, plate xiii, figs. la, b. This species is of rare occurrence in the beds at Glen Rose. It is not a Natica, but inasmuch as only casts are preserved its generic position cannot otherwise be stated. Viviparus (Natica ?) cossatotensia Hill. Arkansas Geological Survey, Annual Report 1888, vol. ir, p. 130, plate iii, figs. 4, 4, 5, 5a. This form was originally described in my Arkansas Report from an isolated locality in Sevier county. Many consider it a Natica, but it is too imperfect to assert its generic position with certainty. The writer inclines to the belief that it is a fresh or brackish water form. Descriptions of Species. 33 Cylindrites (?) sp. indet. The writer has collected from the hase of the Glen Rose beds in Parker county a few specimens of a small gasteropod very much resembling Cylindrites bulletins Lye. and Mort. Unfortu- nately the specimens have been lost. Buccinopsis (?) parryi Conrad. Plate VI, Fig. 1. U. S. and Mexican Boundary Report, p. 158, plate xiii, figs. 4a, b. B. Qmradi Hill. Arkansas Geological Survey, Annual Report 1888, vol. ir, p. 130, plate iii, figs. 2, 2a. Conrad describes this species as follows : ' " Subpyriform ; longitudinally undulated and ornamented with rugose, revolving lines, spire scalariform ; aperture large and patulous." Conrad also says : " Under this name I have described a cast which cannot be referred to any known genus ; it is certainly not a true Buccinopsis, nor can its generic char- acter be determined because of poor preservation." The writer has found casts of this form in great .abundance in the lower fossiliferous horizons of the Glen Rose bed at gypsum bluffs of the Little Missouri, in Arkansas, and at Glen Rose. In some instances the casts show marks in the shape of depressions extending across the lower whorl and the sutures are very deep and distinct. The whorls and spire of a more perfect specimen, shown in plate vi, fig. 1, are six in number and very flat and angular at their upper edge. Whatever may be the true generic position of this form, it is a very useful stratigraphic aid to the student of the Glen Rose beds. The form is readily dis- tinguishable from Tylostonia pedernalis Roemer by its greater number of whorls and entirely different outline of the aperture. Tylostoma pedernalis (Roemer). Plate VI, Fig. 2. Natica pedernalis and N. prxgrandis Roemer. Kreidebildungen von Texas, p. 43, plate iv, figs, la, b'; plate iv, figs, la, b. " Shell large, ovate ; whorls flat, angular above ; lowest or basal whorl three times as high as the spire; spire composed of three coils. Aperture, elongate ovate." Roemer 5 BIOL. Soc. WASH., VOL. VIII, 1893, 34 Hill Paleontology of the Trinity Division. Dr. Roemer described this large and abundant species of the Glen Rose beds from specimens collected 011 the Pedernalis river. He also described Natica ( TylostomcC) prsegrandfy which he asserted differed only by its larger size. The species occurs in the greatest abundance throughout the extent of the Glen Rose beds in Arkansas, Texas, and Mexico, and is preserved only as casts, without shell structure. From observations of hundreds of individuals, I am of the opinion that the two species of Roemer are identical. These forms are indistinguishable from many species described from the Neocornian of France and Spain, where the genus Ty- lostoma has its culmination, ranging, according to Zittel, from the Upper Jurassic. Tylostoma similimum Choffat, from Portugal, a characteristic Neocomian form, is quite closely related, if not identical with the Texas species. It also resembles N. gigcis Bonn, of the Jurassac. The specimen figured here is a large flattened individual, illustrated in order to show the size and variation in comparison with the typical forms figured by Roemer. This form occurs throughout the Neocomian of Mexico, in Venezuela, Bogota, and on the central Pacific coast of South America, and is one of the distinguishing fossils of the Glen Rose beds. Vicarya branneri sp. nov. Plate V, Figs. 1-7. Pleurocera strombiformis (Schlotheim) Hill. Neozoic Geology of Southwest Arkansas, p. 129, plate ii, figs. 1, 2, 3, 4, 5, 6, 7, 7a, 76, 8, 9, 10, 11, 12, 12a; plate iii, figs. 6, 6, 66, fie, copies of European figures. Compare Vicarya helvetica Verneuil and De Loriere. Description de,s Fossiles du Neocomien Superieur de Utrillas et ses Environs, p. 2, plate i. fig. 1. Vicarya lujani Verneuil and De Loriere. Ibid., p. 5, plate i, fig. 3. Vicarya strombiformis Verneuil and De Loriere. Ibid., p. 7, plate i, fig. 4. Vicarya pradoi Verneuil and De Loriere. Ibid., p. 10, plate i, fig. 5. Descriptions of Species. 35 Ttirritella helvetica Pictet and Renevier. Muter, pour la paleont. Suissc du ter. aptien, 1854. ' Cassl.ope helvetica H. Coquand. Monogr. paleont. de Pet. apticn. de PEspagne, 1806. Cerithium lujani Verneuil. Bull, de la Soc. Geol. de France, 2e liv. Tome x. 1853. Cerithmm luxani Verneuil. Memoria gcognostica de -Castellon par Vilanora, 1859, plate iii, Fig. 7. Cassiope verneuilli H. Coquand. Monogr. paleont. de Pet. apt. de PEspagne, 1866. Mui-cf strombiformis Schlotheim. 1820, Petrefact., p. 144. Mui'icitcts strombi/ormis Schlotheim. Monographic der Nord- deutschen Wealdenbildung, Danker, 1846, p. 50, pi. x, fig. 18a, b. This form is indistinguishable from the figures of V. lujani and V. helvetica of De Verneuil and De Loriere, but without compari- son of type specimens their identity cannot be positively as- serted. V. lujani is described as follows : "Shell elongated, with thick test; spire regular, straight, or sometimes a little pupoid; ten or twelve turns of the spire. Each turn of the spire is marked by numerous sinuous cross- stria), strongly bent inward upon the first anterior quarter of each spiral. The sutures are generally well defined." The whorls are also usually marked by two longitudinal elevated bands, one near each border, which in the apical whorls and adult specimens appear as plain elevated bands, or may be in the lower two-thirds of the shell nodular. They continue upon the buccal face or base of the shell as bands. " The lip always presents at the posterior or upper end a pronounced gutter. The outer lip has a deep, broad indentation corresponding with the termination of the basal suture line beneath the last carina or row of tubercules. The inner margin of the labial opening is thickly encrusted." De Verneuil. This is one of the most abundant, variable, and characteristic forms of the Trinity Division, occurring in the lowest molluscan horizons of the Glen Rose beds throughout its extent from An- toine, Arkansas, to the Colorado river in Texas. It was first figured from America by the writer in his report upon the Cre- taceous beds of Arkansas under the name Pleurocera strombi- J'onuls Schloth., after Zittel. Although a very abundant form, the oral aperture was only recently discovered, it having hitherto been broken in the delicate structure of the specimens, and the generic position thereby made uncertain, as is attested by the widely different genera to which it has been referred in Europe. 36 Hill Paleontology of Ike Trinity Division. Tire smallest and largest forms are void of the handsome rows of tuhercules which distinguish the specimens of medium si/e. The larger adult specimens sometimes attain a length of two inches. This form is of interest because it is the characteristic species of the Wealden (Lower Neocomian) beds of Europe. It may be the same as Melania strombiformis, first described from the Wealden strata of North Germany by Schlotheim. De Verneuil and De Loriere, in 1886, published most excel- lent figures and descriptions in their paper upon Materieux pour le Paleontologie de 1'Espagne, entitled " Description des Fossiles du Neocomien Superieur de Utrillas et ses Environs," Paris, 1868. They review the literature of the species and refer it to the Vicai*ya, a subgenus of Cerithidte. They make four distinct species of their specimens, which I believe to be variations of the same species, all of which except one occur in intimate asso- ciation in the lower Glen Rose beds. Professor Jules Marcou, in the previously mentioned review of my Arkansas species, asserts that the form is a Nermasa, but the forms are absolutely void of the characteristic folds which occur upon the columella of that genus, and hence he is mis- taken. The form occurs in great abundance at the gypsum bluffs of the Little Missouri, in Arkansas. At the plant bed locality near Glen Rose, Texas, it is still more abundant and shows the variety helvetica and lujani preserved together in great masses. At Post Mountain, near Burnet, Texas, the badly worn shells of this species occur in an agglomeration ten feet thick (plate v, fig. 7), void of other species and embedded in a matrix of the min- eral grahamite. In this mass all the varieties can be found in association. Nerinaea austineiisis Roemer. Roemer. Paleont. Abhandl., vol. iv, p. 295, plate 31. fig. 8. Fragments resembling this species are abundant in the upper or Mount Bonnel beds of the Glen Rose beds, but are so poorly preserved as to render their assignment to it only provisional. I have found them in the Strontionite beds of the Colorado sec- tion, and a stratum of the beds near the summit of Mount Bon- nel consists almost entirely of calcified Xcruicas. In outer Descriptions of 'Species. 37 marking they correspond to the N. dustinensis of Roenier, origi- nally collected from the Caprina limestone near Austin. The Ncriiieas do iiot'occur in the Comanche series higher than the Caprina limestone of the Fredericksburg Division and not at all in the Upper Cretaceous, where Roemer erroneously supposed this species (originally collected by Mr. Stolley) to occur. Neritina sp. indet. A characteristic fossil of the beds at the base of the Glen Rose subdivision is a small Neritina. Unfortunately the writer's speci- mens were lost in the removal of his collections to Washington, but there are others in the United States National Museum, which he collected from Hood county for the United States Geo- logical Survey in 1886, and it is hoped that they will be figured and described. This form occurs throughout the basement beds in Arkansas and in Hood and Parker counties, Texas. Neumayria walcotti Hill. Plate VIII, Figs. 1, 2, 3. Ammonites walcotti Hill (not Sowerby). Annual Report Ge- ological Survey of Arkansas 1888, vol. u, p. 139, plate i, figs. 1, la, U. Nikitin (Mem. de 1'Acad. St. Petersburg, 1881), defines this genus as follows : Shell flat, widely umbilicate; convolutions thinly rounded, marked by fine falcate lines ; lobes and saddles low, slightly incised ; siphonal lobes longer than the first laterals ; the two lateral and accessory lobes little developed. Only one specimen of this species has thus far been discovered. It occurred in association with 0. franklini, Vycaria lujani, Eriphyla arkansaensis, and other mollusks herein described. The form very much resembles in outward appearance the fig- ures of the genus Oxynotiaras of Hyatt, as given by Zittel and Steinman in their Manuals, but Professor Hyatt refers to it to Neumayria, and contributes the following comments upon the specimen : " Your Ammonites walcotti is probably a Neumayria. The as-' pect is Jurassic, but this group, Upper Jura, and the species 38 Hill Paleontology of the Tr'uutij Dirixiou. nearest walcotti occurs in the very top of the Jura of Central Volga stage, supposed by some to be similar to the Purbeck in the upturn at Malm. The obscuration of a portion of the sutures occurs over the most important part of the outer side, and the structure of the abdomen, which is rounded and has no keel, is not very consistent with the reference either to the \wuim i/riti. of the Jura or the so-called Neiimayr'ta of the Cretaceous. Never- theless it agrees better with those of the Jura than the Cretaceous ones referred to the same genus by Nikitin." Whatever may be the range of this genus in Europe, the writer is inclined to the belief, from the stratigraphy and association, that its occurrence in Arkansas is lowest Cretaceous, and Pro- fessor Hyatt's opinion serves to strengthen the position of the writer in his reticence in earlier papers in expressing a more definite assignment of the Trinity beds before minutely studying the accompanying faunas. The specimen was collected in the banks of Town creek, one mile southeast of Murfreesboro, Ar- kansas. Named in honor of Mr. C. D. Walcott. Acanthoceras (?) justinae sp. iiov. Plate VII, Figs. 1, 2, and 3. Discoidal thin and flattened in general outline, deeply um- bilicate, marked by numerous simple, gently sinuous ribs extend- ing across the convolution and separated by shorter ribs, which disappear near the middle oj' the convolution ; abdomen oblately rounded ; suture lines not preserved in specimens found ; um- bilicus (fig. 2) narrow, depressed. This species .occurs in a coarse sand in the Trinity beds of Travis Peak post-office, in western Travis county, Texas, only four specimens having thus far been found. It is preserved as a cast, unpropitious for the preservation of the shell structure. It is the only ammonitic form thus far found in the Trinity Divis- ion, excepting Neumayrla walcotti, but is of little value in deter- mining the stratigraphic position of these beds. Professor Hyatt has provisionally referred it to the family Acanthoceratidx, and suggests a resemblance to A. rcnwudii Gabb, from the Cretaceous of California. Our species, however, in the writer's opinion, is quite different in general aspect, being very much more flattened, more rectangular at the dorsum, and pos- sessing wider convolutions. It has some external resemblance also to the genus Hoplites. Descriptions of Species. 39 ARTHROPOD A. Cypridea texana sp. nov. Plate I, Fig? ka, 3b. f\ Shell microscopic ; elongate ovate globose, with angular ap- pendage shown in fig. 1 ; opalescent or horny; cardinal margin toothless, thickened centrally ; opposite or ventral margin slightly undulate; surface smooth, and not visibly punctate. It is difficult, says Zittel, to classify with certainty even the families of the Ostracoda, owing to the fact that the shell only is preserved, which is not sufficiently differentiated to justify exact diagnosis. The species here given occurs in abundance in the Glen Rose beds, but usually only the merest outlines are pre- served, or mere granules which suggest the form. In the mol- luscan fauna at the plant bed of the Paluxy, near Glen Rose, however, I was so fortunate as to secure a small fragment in which the shell- structure was well preserved. 'That the species belongs to the Cypridte is strongly probable, resembling both the genera Cypris and Cypridea. I have placed it in the latter pro- visionally, because it is a marine form, occurring in masses of marine shells or mollusca, while the former genus is a fresh water one and of more recent occurrence. There are large masses of sub-oolitic material in the Mount Bonnel beds, which are apparently largely composed of these minute Crustacea. The Cypridse are also abundant in the Lower Neocomian, or Wealden, of Europe. PLANTVE. Undetermined species. Plate I, Figs. 1. a, b, c, d. (" Goniolina?" of author's previous writings.) Spherical cone-like bodies, varying in size from three-quarters to one and one-half inches in diameter ; slightly elongate, oblate or depressed at upper end, with well denned circular scar show- ing attachment to receptacle ; surface consists of minute imbri- cate scales, usually worn down or indistinct ; scales elongate ovate or sub-diamond-shaped, elongated toward upper end, and crowded around receptacular scar ; seed minute. 40 Hill Paleontology of the Trinity Division. This problematic organism has heen provisionally referred to Goniolina in my previous papers. It occurs* from the base to the top of the Glen Rose beds as small spherical calca- reous casts, and extends into the lower layers of the Comanche Peak group at Mount Barker, Travis county. The biologic rela- tions of this organism have been a problem for years, and it has been referred to the Echinodermata, the Foraminifera, and to the vegetable kingdom by various persons to whom it has been submitted. Its occurrence in the chalky strata of the Colorado section remote from other land debris and in association with For-aminifera (Orbitulites texana Roemer) seemed to oppose the fact that it was a fruit or a land plant. The recent discovery by Mr. J. W. Harvey of other plants of many species in the chalky limestone beds near Glen Rose, which have recently been de- scribed in the proceedings of the United States National Museum by Professor Fontaine, dispelled the foregoing hypothesis. Im- mediately beneath the stratum containing the plant bed is another containing many flattened moulds of what could be mis- taken for fucoid stems, and associated with these are numerous specimens of the fossil here figured. A careful study in situ of the surface of a stratum in which these stems were well exposed showed that they branched very much like coniferous plants. At the termination of each ramification was found one of the small spherical casts, as if the limb of a plant laden with cones had been buried in the mud and its cast preserved. Recently, however, the fruit structure has been determined in the speci- mens themselves as figured on plate i. The species should be named for Professor Lester F. Ward, who has done so much for American paleo-botany and has ever encouraged the writer in his studies. The form occurs from Glen Rose southward to the Colorado in great quantities and ranges throughout the Colorado River section. It could be doubtfully referred to the genus Araucarites, which it more closely resembles than any other, although this is for the botanists to determine. This genus is abundantly represented in the Wealden (Lower Neocomian) of Europe and in the Poto- mac formation of this country, as described by Professors Ward and Fontaine. PLATE I. FIGURE 1. An undetermined plant. \ri. Imbricate scales of cone (?), enlarged. \b. Imbricate scales (worn) showing seeds (?). Ic. Scars, apparently attachment of seeds. Id. Outlines of scales - and seeds (?) on worn surface. FIGURE 2. Pattellina texana (Roemer). (Copy of Roomer's figure of Orbitulites texana.) 2". Cross-section, showing irregular structure of interior. 2b. Portion of base, showing irregular granular structure. 2c. Pores of upper surface, greatly magnified.. 2d. Labyrinthoid pores of another part of upper surface, mag- nified. FIGURE 3. Cypridea texana sp. nov. 3. Ordinary aspect of shell. 36. Ventral border of union of valves. All magnified 100 times. FIGURE 4, 4, 46. Serpula paluxiensis sp. nov. FIGURE 5. Anomia texana sp. nov. FIGURE 6. Ostrea franklini ragsdalei var. nov. FIGURES 7, 8. Leda harveyi sp. nov. Magnified 20 diameters. FIGURES 9, 10. Chione'(?) decepta sp. nov. TKOC. BIOL. Soc. \VASH., VOL. VIII, 1893. PLATK I. 10 INVERTEBRATE* OF THE TRINITY DIVISION HILL. PLATE II. FIGURES 1, 2. Pholadomya knowltoni sp. nov. FIGURE 3. Pecten stantoni sp. nov. 3". Detail of marking of larger valve. FIGURE 4. Isocardia (?) medialis (Conrad). FIGURE 5. Distorted specimen. PJIOC. BIOL. Roc. WASH., VOL. VIII, 189:1. J'LATK II. wimw0'^m INVERTEBRATES OF THE TRINITY DIVISION HILL. PLATE III. FIGURES 1, 2. Cucullsea comanchensis sp. nov. FIGURES 3, 5. Trigonia stolleyi sp. nov. FIGURE 4. Trigonia crenulata Roemer. FIGURE 6 Isocardia (?) medialis (Conrad). PKOC. BroL. Soc. WASH., VOL. VI TI, 189.'). PLATE III. INVERTEBRATES OF THE TRINITY DIVISION HILL PLATE IV. FIGURES 1 , 2. Pleuromya (?) henselli sp. nov. FIGURE 3. Pholadomya (?) lerchi sp. nov. FIGURES 4-6. Eriphyla pikensis Hill, magnified 2 diameters. PROC. BIOL. Soc. WASH., VOL. VII F, 180:5. I'LATK IV. INVERTEBRATES OF THE TRINITY DIVISION HILL PLATE V. FIGURES 1-4. Yicarya branneri sp. nov., showing marking of younger forms (V. hijani variety. Magnified 2 diameters V FIGURES 5. 6. Vicarya branneri (V. helvetica variety). Adult form, showing aperture. FIGURE 7. Vicarya branneri. Mass, showing occurrence of species and variations. FIGURES 8-10. Modiola branneri sp. nov. Magnified 2 diameters. PROC. BIOL. Soc. WASH., VOL. VIII, 1893. PLATE V. INVERTEBRATE,-; OF THE TRINITY DIVISION HILL. PLATE VI. FIGURE 1. Cast of Buccinopsis (?) parryi Conrad, showing spire. FIGURE 2. Tylostoma pedernalis (Roemer). Distorted specimen. PROC. BIOL. Soc. WASH., VOL. VIII, 180.' PL ATM VT. INVERTEBRATES OF THE TRINITY DIVISION HILL. PLATE VII. FIGURES 1 and 2. Acanthoceras (?) justina? sp. nov. PROC. BIOL. Soc. WASH., VOL. VI II, 1S measurements here tabulated were all taken from the fresh specimens before skinning, and, unless otherwise stated, the writer is responsible for their accu- racy. So many names have been proposed for white-footed mice from eastern North America that it may appear somewhat haz- ardous to institute still another; hence the species described by authors from the region of importance in the present con- nection may well be considered here in some detail. From the Eastern United States. 59 The first is, of course, the Mas ayrarius atncricanus Kerr (An. Kingd., i, 1792, 231, based on Pennant, History of Quadrupeds, " No. 302 B ")* The description given by Pennant makes special reference to the mixed " dusky and ferruginous " color of the back and ; ' orange coloured " sides of his American Field. Rat, terms which refer unequivocally to our smaller and better- known animal. Any doubt in the case is dispelled by the addi- tion by Pennant in the Arctic Zoology (i, 1784, 131), "length, about four and a half inches ; of tail, four inches ; " thus show- ing that it was the short-tailed form that he had in mind. On Pennant's animal was based also the Mtis sylvaticus novcbomcensis of Fischer (Synopsis Mammalium, 1829, 318), the habitat of which is given as u in Novo Eboraco," and in all proba- bility the Mas noveboracensis of Selys Longchamps (Etudes d'- Micromammalogie, 1839, 67), since this author remarks that the animal is a good species, although considered merely a variety by previous writers. That it is clearly the short-tailed animal that Selys Longchamps refers to is shown by the following extracts from the original description : " Son pelage est d'un fauve plus vif sur les cotes de la tete et du corps. * " * * Longeur totale, 6 pouces 2 lignes ; du corps, 3 pouces 6 lignes ; de la queue, 2 pouces 8 lignes." This mouse is said to replace in North America the European Mas sylvaticus. Rafinesque's Mascidas leucopus (American Monthly Magazine, in, 1818,446) is named among the ten new species of "wild rats " met with by that prolific describer of species during " a journey through the western region of the United States" that is. in the Ohio valley and the pine barrens of Kentucky. As there is little chance that the range of Sitomys canadensis extends to that region, the name is hardly worth considering here. It may be mentioned, however, that Rafinesque's animal is said to be " fallow above," an expression which might apply fairly well to S. americanus, though hardly to the larger form. The next name to be considered is the Cricetus myoides of Gapper (Zoolog. Journ., v, 1830, 204, pi. x)* This animal, from the region between York and Lake Simcoe, Canada, is described as having the " upper half of the body mixed black and light red- dish or yellowish brown." It is further stated that " it measures 31 inches from the tip of the nose to- the insertion of the tail.; the * Synopsis of Quadrupeds, 1771, p. 303, No. 320A (American Field Mouse). History of Quadrupeds, n, 1781, p. 444, No. 302A, 60 Miller Description of a New White-Footed Mouse tail itself, 3} inches." Thus color and measurements alike refer to S. americanus. Moreover, two white-footed mice kindly sent me in the flesh by Mr. I. R. Bourchier, of Button, West Ontario, Canada, just south of Lake Sirncoe, are perfectly typical of the smaller form. Arvicola emmonsii De Kay, from Massachusetts (in Emmons' Report on the Quadrupeds of Massachusetts) is clearly a syno- nym of americanus. The color is given as simply " brown above, darker along the back than the sides," but the whole length is stated to be 6 inches ; tail, 2.5 inches. This animal is said to inhabit " meadows and wooded places. It is often seen in fields recently moived, and is known by the name of Deer Mouse " (italics mine). Sitomys americanus canadensis never occurs in fields and meadows, where, however, S. americanus is often found. Wagner's Hesperomys maniculatus (Wiegmann's Archiv., XT, 1845, Bd. i, 148*), from the Moravian settlements in Labrador, is described as " supra fuliginoso brunneus * * * Korper 3" 2'", Schwanz 2" 5'"." In Beitrage zur Kentniss der Saugthiere Amerikas (Abhandl. Ak. Wiss. Wien, 1848, 315, 316) the author gives practically the same diagnosis, followed by the remarks : " Gestalt, Grosse und Farbenvertheilung verhiilt sich wie bei H. leucopus,.tio dass ich nur die Differenzen anzugeben brauche, welche sich zwischen ihr und dem letztern, von dem ich der- malen nur Beschreibungen, und zwar zumichst die Richardson's vergleichen kann, ergeben. Diese Abweichungen bestehen darin , dass bei H. maniculatus die Oberseite weit triiber gefiirbt ist, indem sie namlich bios russig gelblichbraun und schwarz ge- sprenkelt ist, ohne Beimischung von Rostroth vie es von H. leucopus angegeben wird." This description is somewhat puz- zling, and without specimens from the coast of Labrador it is impossible to decide just what animal it refers to. That S. cana- densis is not Wagner's animal is shown by the measurements, which being taken from " 2 Weingeist Examplaren " must be fairly accurate. Hesperomys campestris Le Conte, from New Jersey, is described so vaguely (Proc. Ac. Nat. Sci. Phila., vi, 1853, 413) that, to use Professor Baird's words (Mam. N. Am., 1857, 485) : " Of the affini- ties of this animal I will hazard no conjecture." That it is not the same as S. canadensis is shown by the measurements length, *Not "1843, n, 141, and 1845, n. 148," as given by Baird and COLICS, From the Eastern United States. 61 3.4 inches ; tail, 2.7 " which were taken from an alcoholic spec- imen. The next name to be examined is the Hesperomys gracilis of Le Conte (Proc. Acad. Nat. Sci. Phila., vn, 1855, 442). Le Conte states that the animal " inhabits Michigan; Professor Baird." From the description, " dark slate color above, a little tipped with brown," it seems probable that the type specimen was immature. There are some discrepancies between the measure- ments given in the original description and those of the same specimen given by Baird (Mam. N. Am., 1857, 473). Accord- ing to Le Conte, the length (head and body) is 3.8 inches ; tail, 4, while Baird gives the dimensions of the same parts as 3.60 and 3.70 inches respectively. The latter author adds that the body is stretched. The long tail of this specimen might suggest the pos- sibility of its being the same as my canadensis, but on account of its poor condition and the vagueness of the description, it seems wisest to discard the name entirely as undeterminable. Moreover, Baird states that the pencil of the type is 0.10 in. (2.5 mm.) in length, which is much less than in any specimens of S. canadensis that I have seen. While not wishing to enter here into a general discussion of the relationships of Mm michiganensis Aud. and Bach., and Mus bairdii Hoy and Kennicott, a few words concerning the probable bear- ing of these animals on the present case may not be out of place. Mas michiganensis, from Erie county, Michigan, is described (Journ. Acad. Nat. Sci. Phila, VIII, pt. n, 1842, 304) as 'a "mouse with yellow cheeks, a light grayish-brown color above, whitish below. * * * The feet, nails, ears, and tail are light brown." It is farther remarked that " there is no dis- tinct line of demarkation. between the colors of the back and under surface, nor does the white extend along the sides as in the white-footed mouse. Dimensions : length of head and body, 4 inches lines ; length of tail, 2 inches 6 lines." That this animal cannot be the same as S. canadensis is shown by the short tail and peculiar coloring of the feet and sides. The description of Mus bairdii (Rep. Com. Patents for the year 1856, published in 1857, p. 92) from northern Illinois and south- ern Wisconsin refers to some short-tailed, bright-colored mouse, quite unlike Sitomys a. canadensis, as the following extracts will show : " Length of the adult male, from nose to root of tail, 2i inches ; tail (vertebrae), II inches ; hind foot, I of an inch. Head 62 Miller Description of a New White-Footed Mouse and body of a large male, 31 inches ; tail, 2 inches. In another specimen, the head and body 3f\ inches; tail, If inches. In spring the hairs of the upper parts are plumbeous at the base, tipped with ashy and yellowish brown ; a few longer hairs, en- tirely black, interspersed. The tips of most of the hairs deepen into black along the back, giving a broad, black stripe when the hair lies flat. In some specimens this stripe is not so dark as in others, but is quite distinct in all, while in some it is pitch- black." It will be remembered that one of the noticeable color features of S. canadensis is the indistinctness-of the dark dorsal stripe; hence Mus bairdii, whatever it really may be, is a very different species. The animal from Burlington,Vermont, described by Baird under the name of Hesperomys myoides (Gapper) (Mam. N. Am., 1857, p. 472), is, in part at least, the same as the subject of the present paper. Baird remarks that " all the white-footed mice from near Burlington, Vermont, had much longer tails in proportion than those from Middleboro, Massachusetts." The only specimens, three in number, that I have seen from the locality in question are, however, typical americanus. Baird 's statement, " tail verte- brae generally .25 of an inch longer than head, and body with a de- cided pencil at the end," and also table of measurements on page 473, refer.without question, to the long tailed form; but his descrip- tion leaves a slight doubt as to just what animal he had in hand. I have never seen a specimen of S. canadensis in which the color is " more vivid yellowish brown " than in S. americanus, nor do any resemble S. aureolus in color, as is said to be the case with U H. myoides," Baird considered the presence of cheek pouches to be the best diagnostic character of myoides. More recently, however, it has been shown by Allen (Bull. M. C. Z., I, 1869, 229) that these structures occur also in the common S. americanm. It is worthy of remark, in this connection, that I have found the cheek pouches of S. canadensis much the more frequently and conspicuously distended with food. Sitomys americanus canadensis is exclusively a Canadian form, replacing S. americanus in the spruce forests of New Brunswick (Restigouche county, E. A. Bangs ; Northumberland county, G. S. Mille;-), and extending south among the hills and mountains at least to central New York and western Massachusetts. Sitomy* americanus is found as far north as Digby. Nova Scotia, and Lake Simcoe, Ontario. Thus the ranges of the two forms overlap From the Eastern United States. 63 geographically about two hundred miles. Nevertheless, the conditions under which the animals live are essentially different, S. canadensis confining itself to dense, preferably damp woods such as Troglodytes hiemalis and Certhia familiar is americana choose to breed in while S. americanus is a mouse of the open fields, clearings, and neighborhood of houses. Only in the central and southern part of its range, where the character of the country is very -different from that inhabited by S. canadensis, does the smaller animal take to the woods with anything like regularity. I have no doubt that the northward range of S. americanus has been considerably extended by a gradual movement, following the clearing away of the forests, thus bringing the two races into their now curiously close juxtaposition. For the present at least I have thought it best to treat these two animals as subspecies. It must be confessed, however, that the number of intermediates is surprisingly small, less than a dozen in the total number of specimens examined, and that these occur in no particular geographical region. The case is suscep- tible of no definite proof until more facts are forthcoming ; meanwhile it lies with each. observer to treat these closely allied forms as his individual preference may dictate. 9-Bioi,. Soc. WASH., VOL. VIII, 1893, 64 Miller Description of a New White-Footed Mouse Measurements of One Hundred and Fifty Specimens of Sitomys americanus (Kerr}. Number. Locality. Date. M A .d "fc/D p " .1 Tail vertebra'. TJ 1 1 13 3- ^ u 1 & If "^ z '"<= K~ oil? -. -_ Skin. Skull. 2011 1763 Sutton, West On- Nov. 25, '92 9 ir>7 74 3.8 21.6 17 443 tario, Canada. V . 2012 1764 It It Dec. 1,'92 9 160 70 3.6 20 IT). 4 43.7 2231 Pigby, N. S ..... ( )ct 9 '92 3 166.5 SO. 5 3.4 20.5 17.5 47.7- 2232 9 '92 , 164 7S 2.8 20.5 It; 47.5 2233 " 11 '92 V 1 79. 83 3 19.5 17 4X.9 2234 ' 13, '92 167.5 K2 4 19.5 15.5 4S.3 2235 ' 13, '92 9 ! 166.5 79 3.8 20 15 47.4 2236 < 1(3, '92 i <~T 163 79 3 19 17.5 48.4 2237 ' 16' '92 9 156. 5 ! 72.5 2.4 20 15.5 46.3 2238 ' 17 '92 Q 154.5 70 -S .'5.4 20.5 Hi 44.9 2239 ' 17, '92 $ 156.5 74.5 3 ><>.- Hi 41.2 2240 " 21, '92 1 Q 157 71 3 19 17 45.2 2241 " 23, '92 9 166.5 80 3.4 20.5 14.o 4S.4 2242 " 2;}, '92 c? 159 73 2.4 19.5 15.5 45.9 1024 873 Elizabetb town, Dec. 20, '91 \J *v/ 9 169 74 3.6 20.4 17 j 43.1 N. Y. 1025 874 U 11 " 21, '91 ^ 165 74 3.4 22 16.8 44.8 1042 890 t (I " 27, '91 r? 170 76 4 20 16.4 44.7 1049 897 I 11 " 28, '91 9 180 79 3.2 19.5 16.2 43.9 1054 902 I 11 " 29, '91 cT 167 68 3.2 20.4 16-2 40.7 1055 903 t 11 " 30, '91 cT 162 68 3.6 20.4 16.2 41.9 1059 907 I It " 31, '91 168 74.8 3.6 21 14.8 44 1060 908 ( " 31, '91 9 160 71 3.4 20.5 15.4 44.3 1063 911 ( Jan. 1,'92 ^ 168 71 4.4 20 14 42.:', 1065 913 <( " 1, '92 c? 158 66.5 3 20.2 148 42.1 1066 914 U tl " 1, '92 9 158 70 3 20 15.6 44.3 1140 967 11 tl " 3, '92 181 83 4.2 21 17 .45.8 1141 968 it a " 3, '92 ' cT 169 71 3.4 21 17 44.1 1282 1109 It tl Feb. 27, '92 c? 166 74 3.6 21 16 44.6 1292 1112 It 11 Mar. 10, '92 154 69 2.8 20.5 15.5 44.8 1352 1169 (( 11 Apr. 3, '92 9 163 68 3.2 20 16 41.7 1353 1170 It It " 4, '92 (j) 157 67 3 20 14.8 42.6 1355 1172 ti It " 11, '92 9 172 71 3.8 19 15.2 41.3 1356 1173 tl tl " 4, '92 Jp 176 80 4 20 14.5 45.4 1357 1174 tl It " 5, '92 ^ 175 76 2.8 19.4 16.8 4:5.4 258 . Peterboro, N. Y. July 15, '90 9 175 76 3.6 19.3 14 43.5 1595 'i392 tt it " 20, '92 9 192 86 4.4 21 16 45 1640 1437 tt it Aug. 3, '92 9 172 76 .... 20 16.4 44 1648 1444 ti n " 17, '92 o 172 76 4 19 17 44 1649 1445 ti 92 9 155 69 2 19 17 45 1709 1500 " 17, '92 $ 156 71 2 20 16 45 1710 1501 " 17, '92 j 167 76 3.4 20 17 46 1716 1507 . ii " 25, '92 ; c? I 152 66 3.4 19.8 | 14.4 , 43.4 * Collected and measured by Outram Bangs. From the Eastern United States. 65 Measurements of One Hundred and Fifty Specimens of Sitomi/s americanus (Kcrr). ^ a Number. \ i 1 +J | >o Locality. Date. ^ 1 1 | f 3 . Skin. Skull. M 3 o 1 o q 0> -d S Ill GO H H PH K H *" 1997 1760 Peterboro, N. Y. Nov. 9, '92 c? 168 75 3 22 16 44.6 283 Geneva N Y Dec. 25, '90 9 165 71 3.4 20.4 17 43 284 ti it " 27, '90 9 165 72.5 3.8 19 16 43.3 1163 "980" West Dedluini, Jan. 30, '92 c? 184 82 3 20.4 14.5 43.5 Mass. 1154 981 ti it " 30, '92 cT 178 78 3.2 20.8 15 43.8 1155 982 n n " 30, '92 ^ 163 69.6 3.2 20.5 16 42.7 1156 983 " 30, '92 $ 169 72 3.2 21 15.8 42.6 1157 984 " " 30, '92 <~p 183 82 4 20 16 44.8 1158 985 " 30, '92 c? 165 70 3 21 15.8 42.4 1159 986 " 30, '92 179 78 3.4 21 14.8 43.6 1160 987 ' 30, '92 9 172 79 4 21 15.2 45.9 1161 988 " 30, '92 9 165 72.5 4 20.5 15 43.9 1162 989 " 30, '92 9 179 78.5 3.4 23 16.5 43.9 1163 990 " 30, '92 9 173 76 5 20.2 17 43.9 1164 991 ' 31, '92 c? 162 69 3.4 20.5 14.8 42.5 1165 992 ' 31, '92 167 76 3.6 20 15.4 45.5 1166 993 ' ' ( 151, '92 9 163.5 72 3.2 19 13.5 44.3 1167 994 ' < ' 31, '92 9 163 72 4 21 17 44.1 1168 995 ' ' ' 31, '92 9 163 70 3.8 20 14.8 42.9 1402 1218 ' ' May 8, '92 9 182 81 2.4 20 16 44.5 1173 1000 ' ( ' " 1,'92 j( 17.1 74 3.4 20.5 15 43.2 1174 1001 ' ' " 1, '92 9 159 66 3.2 21.4 16.4 41.5 1175 1002 11 i " 1. '92 9 174 77 5 20.2 15 44.2 1402 1218 it i " 8, '92 9 182 81 2.4 20 16 44.5 448 372 Seekonk, Mass. . . April 6, '91 9 195 88.5 3.4 20.8 17.2 45.4 919 805 it i Nov. 28, '91 c? 170 79 4.6 21 14.8 44.1 920 806 n i " 28, '91 152 66 20.4 14.8 4^.4 922 808 it i " 29. '91 ^ 167 87 3.4 20.2 15 46.7 923 809 11 t ' 29, '91 ^fr 162 67.4 3.8 20.4 15 41.6 924 810 t: I ' 29, '91 g 184 85.5 2.8 20.8 15.8 46.5 926 812 11 ( ' 27, '91 9 190 88.5 32 20.8 15.6 46.6 927 813 11 I ' 27, '91 9 171 76.6 3 20.5 15.6 44.8 928 814 11 ( ' 29, '91 9 188 84 4.2 20.4 17 44.7 929 815 11 I " 29, '91 9 181 76 4 20.5 15 42 931 817 11 " t " 29, '91 9 196 94 3 20.6 15.4 47.9 932 818 11 I " 29, '91 163 75.6 3.6 20.8 16.4 46.4 9:;:5 819 11 I " 29, '91 9 180 79 3.6 19.8 14.4 43.9 936 822 I. I " 30. '91 6 178 75.5 3.2 19.8 15.8 42.4 2014 11 I Dec. 12, '92 9 *193 86 4.2 20 16 44.6 22S9 Concord, Mass . . . Nov. 5, '92 _7\ 159 69 4 20 15 43.4 2290 it n " 5, '92 cr 158 69 3 20.4 15 43.6 1399 1215 AVeston, Mass. . . . May 2, '92 cT 170 72 3 20.8 17 42.3 1400 1216 i " " .... " 2, '92 g 180 79 2.8 20.8 16.8 43.9 1401 1217 U " 2, '92 9 166 68 2.6 20 17 41 436 437 357 358 Ipswich, Mass . . . Mar. 21. '91 " 21>91 i 176 162 82.5 75 3 2.6 20.5 20 16 15.8 46.9 46.3 439 360 North Truro, Mass. April 1,'91 9 i 188 83.5 3 19.5 16.2 44.4 440 361 " " " 1,'91 9 156 66 3 19.5 17 42.3 66 Miller Description of a New While-Footed Mouse Measurements of One Hundred and Fifty Specimens of Sltomys americanus (Kerr}. Number. Locality. Date. 1 H I Tail vertebrae. Pencil. tj ig Ear from notch. 11 'B-3 'o t,~ .gig 5 |.S.2 Skin. Skull. 441 442 443 444 451 749 1422 1525 1527 1529 1530 1531 1 3 4 5 7 9 10 13 14 15 16 17 18 19 20 1 1 13 15 18 30 32 38 51 57 88 90 118 125 146 152 134 135 58 61 71 362 363 364 365 366 642 1234 1331 1333 1335 1336 1337 North Truro, Mass. u it it tt n it tt a u a u tt a n 11 U U 11 u u Liberty Hill, Conn. t n t a t tt I it t u it ti tt it a a a a tt n a tt tt . a it it a it Haddonfield,-N.J. a a n a a a a it tt a tt a tt tt a tt a tt tt it n a a tt Thorndale, Pa Barren Ridge, Pa. Marple, Pa U U a tt April 1,'91 " 1, '91 " 1,'91 " 1, '91 " 1,'91 Sept. 22, '91 Mav 22, '92 June 24, '92 " 24, '92 " 25, '92 " 28, '92 " 28, '92 Nov. 21, '92 " 22, '92 " 22, '92 " 22, '92 " 23, '92 " 26, '92 " 26, '92 " 26, '92 " 26, '92 " 26, '92 " 26, ? 92 " 26, '92 " 26, '92 " 26, '92 " 26, '92 Dec. 14, '92 " 15, '92 " 24, '91 " 24, '91 " 24, '91 " 26, '91 " 26, '91 " 27, '91 " 28, '91 Jan. 11, '92 " 22, '92 " 23 '92 f Feb. 2, '92 " 3, '92 " 13, '92 " 18, '92 " 7, '92 " 7 '92 Jan. 13/92 " 14, '92 " 15, '92 | tf 1 1 I 1 9 c? 152 148 153 156 1(14 169 182 183 167 188 175 180 178.5 145 173.5 164.5 158 171 167.5 151 146 142 166.5 151 165 158 164 175 173 169 165 141 152 143 155.4 156 153 174 159 171 152 165 157 161 149 152 157 140 95 67 65.5 68 69 78 85 82 74 83 80 79 78 57.5 74.5 70 71 77 78.5 69 66 61 74.5 65 75 69 74 80 73 73 74.7 62 66.5 63.5 71.6 66 64 82 69.8 74.2 67.3 73 67 67 58 63 67 57 4 3.2 3 3.6 2.6 3 2.4 2.4 2.4 3.8 3 4 3.8 3 5.2 3.8 4 3.2 3.6 3.4 4.2 3.8 3.4 3.4 3.4 3.8 5 3 3 4 3.6 3.2 3.6 4 4.2 4.2 3.8 4 4 3.6 3.4 4 3.8 4 2.4 20 20 19-5 20 20 21 21 20.2 20 20 21 20 20 20 20.5 20.5 20.5 21.5 20 19 20 18 19 20 19 18 19.5 20.5 21 21.3 19.3 19.3 20.3 19 20.3 19.3 21.6 20.3 19.6 23 21.3 20 21 21 20.3 21 20.3 20.3 16 lli.4 16 14.5; 16 16.8 17 16.2 17 16 15.4 16.8 17.5 14 15 15 14.5 19 18 18 18. 15 17 17 15 15 17.5 17 15.5 15-2 16.3 J5.7 16.3 14.7 17.5 15.7 , 16 17 16.5 16 10 15.2 15.7 16 15-2 16.8 42.1 45.:; 42.1 43.6 42,1 46.1 46.7 44.8 44.3 44.1 46 44 4:5.2* 40 42.9 41.9 44.2 45 46.9 45.6 45.2 43 44.7 43 45.4 43.6 45.5 45.7 42.2 41. 6f 45.3 44 43.7 43.7 46.1 42.3 41. S 47.1 43.9 S 43.4 44.3 44.2 42.6 : 40.3 39.6 41.4 42.6 40.7 i * Collection of Outram Bangs; measured by collector, f Collection of S. N. Rhoads; measured by collector. From the Eastern United States. 67 Measurements of One Hundred and Fifty Specimens of Sitoinys americanus (Kerr), A , 0) 3 Oi i> Number. fi C 4 *a o -;** Locality. Date. a ! 1 o ^ + '^- Skin. Skull. H 1 t> i 'o 1 5 ft 1 PH 74 Marple, Pa Jan. 16, '92 tf 154 63 3 20.3 16 41 75 " 16 '92 % 156 69 3 8 22 15 9 44 3 2132 I 18irf Washington, D. C. Feb. 12, '93 9 194 91 2 21 17 46.9 2133 1863 tt tt " 12, '93 9 185 85 2.8 20 15 45.9 2134 1864 tt tt " 12, '93 d 1 165 76 3.6 21 16 46.6 2135 1865 n n " 12, '93 cf 177 84 3.4 20 16 47.4 2136 1866 tt a " 12, '93 d^ 165 72 3 21 16 43.6 Measurements of One Hundred Specimens of Sitomys americanus canadensis Miller. 8 1 IJ Number. i E ,O | o . G '3 o Locality. Date. s o> 1 g p offi~' Skin. 'Skull. H i > i 1 o> 1 A M eg e^ OX 1 J3 ,~_a> a> C H H PH 3 N 1436 Northumberland June 6, '92 C? 170 83 19.8 16.8 48 8 Co , N. B. (J TlO. O 1437 a a " 9, '92 9 150 73 19.8 16 48.7 1419 1231 Oak Bay, N. B. . . . Apr. 11, '92 155 75 5 19.8 15.6 48.4^ 1848 1626 " " May 18/92 ^ 185 90 6 20 16 48.6 H'32 881 Elizabeth town, Dec. 23, '91 tf 181 91 7.4 20.2 17.8 50.3 N.Y. 1033 882 a a " 23, '91 ? 197 100.5 8 20 17.4 51 1034 883 i a " 23, '91 ira is a terebratuloid loop having a Centronella-like form, which un- dergoes several modifications before the growth of the spiral lamellae, and thus in so far resembling the development of Dielasma. These results threw doubt on a number of Lower and Upper Silurian species described as having recurved loops and pre- viously referred to the higher terebratuloid genera Mawmlrcria or Waldheimia. The shells are impunctate, while Rertesdaeria and Centronella are distinctly punctate, like all other well-known terebratula?. Upon investigation, it has been ascertained by Hall and Clarke and the authors that the species which have been referred to Hallina and Macandreria from the Silurian are spire-bearing forms, and therefore do not belong to the Ancylo- brachia. Fischer and (Ehlert have called attention to a number of recent species which have been erroneously based upon the immature stages of higher species, and in the Terebratellidse it is evident that great uncertainty must exist in the identification ot specimens not fully adult. Now, finding that Paleozoic genera Brachial Supports in Dielasma and Zygospira. 73 of botli loop and spire-hearing stocks (Ancylobrachia and Heli- copegmata) in the adolescent period likewise pass through metamorphoses representing the structures of other genera and even other suborders, it is manifest that species cannot be re- ferred to their proper genera nor genera correctly defined unless the individuals studied are adult and their characters constant for a definite period of time. Development of the Loop in Dicla-aina turr. 20- The brachial supports in a mature specimen. X 6. The specimens figures 12-20 are from the top of the Trenton, Frankfort, Kentucky. 21. The spirals and loop in a Canadian specimen of (Aimzijya =} Zygospira recurvirostra. (After Davidson.) X 3. 22. The spirals and loop in (Hallina =} Zygospira sajfon!! \V. and S. Trenton, Tennessee. X (>. 23. The spirals and loop in (Hallina =) Zygospira nicoll'tti W. and S. Trenton, Minnesota. X 6. 24. The 'spirals and loop in Zygospira headi Billings. (After Hall.) X 2. 25. The spirals and loop in Zygospira modesta Say. (After Hall.) X 2J. 26. The same. (After Davidson.) X "]. PHOC. BIOL. Soc. WASH.,- VOL. VIII, 1893. 1 2 BKACIIAIL SUPPORTS IN DIELASMA AND ZYGOSPIHA. ON THE DEVELOPMENT OF THE SHELL OF ZYGOS- PIRA RECURV1ROSTRA. BY CHARLES SCHUCHERT. The material showing the ontogeny of the shell in Z. recurvi- rostra was gathered from some blue-green shales on St. Anthony hill, a suburb of St. Paul, Minnesota. This horizon is equivalent to the lower Trenton of New York. Some of the associated brachiopods are Orthis testudinaria, 0. meedsi, Plectambonites sericea, Strophomena scqfieldi, Clitambonites diversa, etc. The youngest specimen observed (plate xi, fig. 1) has a length of .8 mm., and is elongate subtriangular in outline, biconvex, with the ventral valve a little the deeper ; ventral beak acumi- nate, inclined posteriorly at an angle of about 45 to the plane of the valves ; delthyrium triangular, as wide as long and devoid of deltidial plates. In the apical portion there is a short concave plate continuous with the walls of the delthyrium, but apparently not attached to the rostral cavity. The fold and sinus are faintly developed, becoming obsolete at about the center of the shell and are without plications. In other individuals of about the same size the sinus is occupied by three short plica- tions and the ventral fold by two. In specimens of a somewhat larger growth these primitive plications are rapidly followed by a number of new ones along the entire anterior margin. The size of the shell at which they begin to develop is variable (com- pare figs. 2-5), being the earliest in the narrow depressed indi- viduals (fig. 3) and latest in the rounder and more convex specimens (fig. 4). New plications are rarely interpolated, their number being increased as growth proceeds by the addition of others along the lateral margins of both valves. The plications remain simple throughout. The smooth nepiastic stage gradually grows more and more rotund and subquadrangular in outline, and at maturity is pli- cated to the apex of the shell. The ventral beak, which is at first slightly recumbent (fig. la), becomes erect, and finally is strongly incurved over the dorsal umbo (fig. 9a). The large 11 BIOL. Soc. WASH., VOL. VIII, 1893. (79) 80 Schuchert On the Development of the open delthyrium is gradually reduced in size by the introduction of deltidial plates which grow inwardly from the walls of the fissure, heing wider anteriorly where they join, leaving in the apex an oval pedicle opening. As the beak incurves these plates become larger, stronger, and anchylosed along the median line, but at maturity are nearly completely hidden by the dorsal limbo. The pedicle opening at maturity (fig. 9) is through the acutely convex portion of the ventral umbo, and is compara- tively smaller in size than during previous stages. Growth Stages. A well-preserved specimen of about 1.5 nun. length (fig. 10), shows three distinct stages of growth before the introduction of the plicated or specific period : a, the initial shell or protegulum, with both cardinal lines arched ; /;, a broad, oval stage, in which the ventral hinge areas on each side of the delthy- rium first appear, followed by c, a subcircular form, with the beginning of the fold and sinus. It is either during stage c or b or both that the concave plate in the apex of the delthyrium is developed. During the next or fourth stage the first specific characters begin to appear, as shown by the plications, and also the first stage of the calcareous brachial supports. Observations and Correlations. The first or initial shell in Zygospira, as in other brachiopods' is the protegulum, which has been compared with adult Pater in a of the Lower Primordial. In many inarticulate brachiopods it is known that the protegulum is followed by a nearly round Obolella-like inarticulate stage, but in all rostrate articulate spe- cies in which the second stage has been observed there appears the first articulation of the valves. The fold and sinus, along with a few rudimentary plications, are introduced during the third stage of Zygospira. This form of shell much resembles some primordial species which have been provisionally referred to Camarella. With but slight modifications in the convexity of the valves and -the greater or less prominence of the fold and sinus, this form is repeated in a number of early Paleozoic genera of the suborders Truttacea and Rostracea, as primordial C'lmarr/lii and many species of P&ntameru8j Zyyoxpira nirolrfti, (.'/oK/>ir. 11, 1 la. An unusually large specimen of (Hallina =) Zygospira nicoletti W. and S., in which a few plications are developed on each side of the fold and sinus. X 6. 12, 12a. Dorsal and profile views of one of the types of (Hallina =) Zygospira nicoletti AV. and S., the form most commonly observed. X 6. 13, 13. Dorsal and profile views of the type of (IlaUhia ) Zygo- spira saffordi W. and S. This species is closely related to Z. recurvirostra. X 0. Pnot 1 . BIOL. Soc. WASK., VOL. VIII, 1893. 123 4 PLATE XI. DEVELOPMENT OF THE SHELL OF ZYGOSPIKA RECURVIROSTRA. VOL. VIII, PP. 83-96 JULY 18, 1893 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON REDISCOVERY OF THE MEXICAN KANGAROO RAT, DIPQDOMYS PHILLIPSI GRAY BY C. HART MERRIAM, M. D. WITH FIELD NOTES BY E. W. NELSON. The first kangaroo rat known to naturalists was described by J. E. Gray, in 1841, under the name Dipodomys phillipxi* Both species and genus were new. and were based on a single speci- men presented to the British Museum by John Phillips on his return from Real del Monte, in the state of Hidalgo, Mexico, in the vicinity of which place it was supposed to have been obtained. f This specimen still remains unique, so far as authen- tic published records go, notwithstanding the fact that numer- ous museums in this country and Europe contain specimens labeled D. phillipsi which have, been examined and treated as * Dipodomys phillipii Gray, Annals and Magazine Nat. Hist.,-vn, Aug., 1841, 522. In the original article the name was. spelled phUlipii by typo grapical error ; it was corrected to phiUipsii a few months later by the same author. (Am. Jour. Sci., XLII, 1842, 335.) t Real del Monte is in the mountains at the extreme north end of the 'Valley of Mexico, about 50 miles northeast of the City of Mexico. 12-Bioi,. Soc. WASH., VOT,. VIIT, 1893. (S3) 84 Merriam Rediscovery of the Mexican KIK/<> Eat true philiipsi by nearly all writers on the genus during the past half century. In studying a large series of kangaroo rats from Texas, New Mexico, Arizona, and California several years ago I was sur- prised to find that none of them conformed with the original description or with Audubon and Bachman's plate and measure- ments taken from the type specimen in the British Museum. It became apparent therefore that true Dipodomys philiipsi was not represented in the extensive collections examined, and probably was not an inhabitant of any part of the United States. Every available means w r as used to secure specimens from the sup- posed type locality, but without success. Letters were written in both English and Spanish to Mexican officials in Real del Monte describing the animal and offering a reward for a speci- men, and later aii experienced mammal collector was sent to the place, but with no better result. Finally Mr. E. W. Nelson. who has caught hundreds of kangaroo rats in various parts of the United States and Mexico and is thoroughly familiar not only with the habits of the animals but also with the kind of country inhabited by them, was requested to visit the region, in connection with his work for the United States Department of Agriculture. The result is a series of 67 handsomely prepared skins and skulls and several complete skeletons of the long lost Dipodomys philiipsi. Hence, after a lapse of more than fifty years, it is now possible for the first time to redefine the type of the genus and to differentiate the species from others with which it has been persistently confounded. Before doing this, however, it seems desirable to put on record with some detail the facts connected with its rediscovery and distribution. Notes on a Search for Duplicate Types of Dipodomys philiipsi. In response to my request, Mr. Nelson has contributed the following interesting account of his search for this species : '" After securing specimens of Dipodomys philiipsi in the Valley of Mex- ico, at Tlalpam, Ajusco, and Amecameca in December, 1892, and February, 1893, I finally set out to try and obtain specimens nearer Real del Monte, the supposed type locality. A long and careful search in the vicinity of Tula, in Hidalgo, just north of the valley, failed to discover it or any allied species. From that point to Pachuca, along the extreme northern border of the valley, the country presented the same hard, rocky, and unsuitable character found at Tula. Pachuca is situated at the extreme Dlpodomys pliilllpsi Gray. 85 northeastern point of the Valley of Mexico, at the base of the Sierra de Paclmca, the latter forming the limit of the valley in this direction. At an altitude of about 9,000 feet, and about three or four miles in a direct line higher up in these mountains, lies the mining town of Real del Monte, the reputed type locality. The mountains are almost wholly composed of porphyry, with only a very scanty coating in places of a hard, clayey soil. A short visit to this locality was enough to show very conclusively that no species of this group had ever been taken in its im- mediate vicinity. A visit to the northern slope of the mountains near El Chico showed it to possess the same character. " The next effort was made in the forlorn hope of finding the species about the border of the valley, at the base of the mountains, near Pachuca. A careful search of this vicinity for some miles in various directions failed to discover any sign of kangaroo rats or, indeed, any sufficient area of suitable ground. A trip was then made to the village of San Augustin, 18 miles south of Pachuca, in the Valley of Mexico, and this also resulted in finding the same rocky hills and hard clayey bot- toms that prevail about the northern end of the valley. " Giving up all hope of finding the species near Pachuca or Real del Monte the alleged type locality I decided to make another search for it at Irolo, which lies a little more than 25 miles south of Pachuca, in Hidalgo, and just east of the low divide which borders the Valley of Mexico on the east between the Sierra de Pachuca and the northern end of the Sierra Nevada de Iztaccihuatl. The course of the Vera Cruz or Mexican railway, on which I traveled, at first kept to the south near the west side of the divide. Some six miles out of Pachuca we passed over a narrow belt of softer soil than usual here, and I had hasty glimpses of several burrows among the maguey plants that I was quite certain be- longed to some species of kangaroo rat. The road then left the divide and swung out into the Valley of Mexico, and the soil again became a hard clay. From the station of Ometusco, just within the Valley of Mexico, the road leads east for a few miles over a low dividing ridge to Irolo. About the latter place and away to the northward, toward the mountains of Pachuca, the bases of the low hills and ridges show pale yellow shades such as sandy deposits would give. A short search about the base of the hills near Irolo showed that areas of sandy and loose soil did in fact exist there. " Further search showed that these areas of soft soil were occupied by kangaroo rats, which when captured, however, proved to be a 5-toed Pfrntlipiiii instead of a 4-toed Dipodomys. No trace of any other species could be found, but the Perodipus was abundant. " From the fact that the belt of country in which this species occurs at Irolo extends directly toward the point where I saw the signs on the road from Pachuca, with less than twenty miles intervening and a divide of not over 250 feet. I feel fully justified in assuming that the animals whose burrows were seen near Pachuca are a colony of the same species as that found at Irolo, namely, a Perodipus. 86 Mcrriam Rediscovery of the Mexican Kangaroo Rat " On the other hand, the city of Tlalpam was the capital of the state of Mexico from 1828 to 1830, and from a much earlier period has been one of the important towns of the Valley of Mexico. This being the case, the abundance of Dipodomys in the sandy land at the. very border of the town suggests the probability that a specimen taken there by an early traveler may have been the one brought to London by Mr. Phillips of Eeal del Monte, the erroneous reference arising from the notorious care- lessness of the early collectors in labeling specimens." - It should be added that Tlalpam is only about 60 miles from Real del Monte, and should be remembered that Gray did not say that his specimen came from Real del Monte, but merely that it was sent to the British Museum by " Mr. John Phillips, who has lately returned from Real del Monte " quite another matter. The probability is exceedingly great that the specimen was not labeled at all before it entered the British Museum. Geograph ic Distribution. Mr. Nelson first found Dipodomys phillipsi at the- extreme southern end of the Valley of Mexico and on the adjacent mountain slopes, where 28 specimens were obtained. Concern- ing its occurrence in this region he writes : "These kangaroo rats occur in the bottom of the valley near Tlalpam, D. F., at an altitude of 7,500 feet, and at the eastern base of the main peak of Ajusco, near the village of the same name, at an altitude of some- thing over 10,000 feet. Still further to the south, on the extreme border of the state of Mexico, adjoining Morelos, they were noted close to the peak of Huitzilac, near the Cruz del Marquez, at an altitude of 9,000 feet. They were also taken at the western base of Mt. Popocatapetl, near the village of Amecameca, at an altitude of about 8,300 feet. The peak of Huitzilac lies about 20 miles south of Tlalpam, and Amecameca is 28 miles east of the peak of Ajusco, thus giving the species a known range of less than 20 by 30 miles in this district. " The animals were far more numerous near Tlalpam than elsewhere. This place is 9 miles south of the City of Mexico, at the extreme south- ern border of the valley, just where the first slopes of the Sierra de *ln continuing my work to the eastward 1 found it necessary to visit, the city of Puebla. There 1 learned that the State College had a small museum, which I visited. One of the first things that met my eye was a D'podomys phillipsi labeled as coming from Ojo de Ayua, Puebla. Leav- ing the city, my route took me first into Tlaxcala and thence to eastern Puebla, in both of which districts I found the species abundant, as de- tailed elsewhere in this paper. Dipodomys phillipsi Gray. 87 Ajusco begin to form the inouiitnin rim on that side of the basin. Lying just west of Tlalpam is a great lava bed, known as the Pcdiryul, which descends from the northern base of the peak of Ajusco and reaches down the slope and out along the southwestern border of the valley. The peak of Ajusco, rising to an altitude of 12,600 feet, lies about 9 miles in a southwesterly direction from Tlalpam. Commencing near the eastern base of this peak and following down the eastern border of the lava bed to the bottom of the valley at Tlalpam is a bed of fine, grayish volcanic ashes or sand. Originally this deposit of sand did not reach the bottom of the valley, but the heavy summer rains gradually washed it down until it is now spread for a mile or two out over the bottom, at the base of the hills extending east from Tlalpam. In this bed of fine, sandy soil, beginning almost with the last houses of the town, these kangaroo rats are very abundant. The ground occupied by them at Ajusco, Htiit- zilac, and Amecameca was once covered with pine timber, but is now used as corn or grain fields, and it is altogether likely that the distribu- tion of the species in all of these localities from the border of the Valley of Mexico followed the subjection of the land to cultivation. At the present time, although there are large areas of apparently suitable land, its distribution is local and restricted to comparatively small districts." At a later date Mr. Nelson found Dipodomys phillipsi on the plains of Puebla and thence northerly into eastern Tlaxcala and western Vera Cruz. Concerning its occurrence in this region he writes : " The species was not encountered again until I reached the state of Tlaxcala. There, in the eastern half of the state, about the northern and eastern base of the Cerrq de Malinche and the towns of Huamantla and San Marcos, it is common. Extending thence easterly along the same sandy plain into the state of Puebla, it is also very numerous. " In this latter state the species reaches its extreme southeastern limit near the towns of Canada Morelos and Esperanza. To the north of these localities it is abundant about the towns of San Adres, Chalchicomula, and up the adjacent western base of Mount Orizaba to an altitude of 9,000 feet. From these points it ranges across to Ojo de Agua and San Juan de los Llanos, in the same state. To the northeast its extreme limit is found about the northern and western base of the "Cofre de Perote, a little east of the town of Perote, in the state of Vera Cruz. This gives the species a known range over parts of four states. It occupies the extreme southern end of the Mexican table-lands, and is therefore the extreme southern representative of the genus. " So far as my observations have extended, this species is confined strictly to the areas of volcanic sand produced by two volcanic centers : first, the sandy belt lying between Mount Popocatapetl and the Cerro de Ajusco, some thirty miles in a westerly direction from the first-named peak, and all within the State of Mexico ; second, the much larger district having its 88 Merriam Rediscovery of the Mes'u-an l\u<> Hat eastern border along the volcanic range from Mount Orizaba to the Cofre de Perote, and extending thence west to the eastern base of Mount Popo- catapetl, crossing a small corner of Vera Cruz and all of the states of Puebla and Tlaxcala. Since the sandy plain of southern Tlaxcala and central and western Puebla as far as the eastern base of Popocatepetl is a continuation of the same character of country as that in which I have found the species in adjacent parts of these states, I have no doubt that further work will show it to be equally common in these unexplored dis- tricts. That the species ranges from Puebla and Tlaxcala northwesterly toward its reputed type locality in the Sierra de Pachuca is improbable, since a high pine-covered area separates the two districts. " The presence of this species in Puebla and Tlaxcala adds considerably to the uncertainty attending the probable source of the type specimen. While Tlalpam was a well-known locality of early days in this country, it is to be taken in consideration that the regular highway (stage road) from Vera Cruz to the City of Mexico passed Perote and traversed a hun- dred miles of country where this species is very common. The vertical range of this species lies between 7,400 and 10,000 feet, but by far the greater number of individuals are found between 7,500 and 8,500 feet. Its center of abundance is in southeastern Tlaxcala and adjacent parts of eastern Puebla. Its extreme upper range, both at Ajusco and Mount Orizaba, is due to the animals following cleared fields up into the pines from their original lower range on the plains." Habits. Respecting the habits of Dipodomys phillipsi Mr. Nelson con- tributes the following, which, unless the contrary is stated, re- lates to the neighborhood of Tlalpam : "Their little trails were to be seen after a calm night crossing the sand in every direction, forming in many places a perfect network. Each animal occupies a burrow having as a rule a single entrance, dug in the bare, open field. The holes enter the ground at a slight angle, and each has a shallow trough-like depression leading out from it for a few inches, as is customary with most members of the group. Ordinarily the trail leaving the burrow forks, forming a Y very close to the entrance, and each branch trail leads away across the sandy soil to a neighboring hole, or to a distance where it becomes lost among the scanty herbage, where the owner finds its forage of seeds and small succulent leaves. Here and there little cone-shaped pits, an inch or two deep, show w r here the inhab- itants have dug up little plants or hidden seeds. " When captured the animals frequently had their cheek pouches full of seeds and small green leaves or young plants. Judging from what I noted in this way, it was evident that quite a variety of small plants contribute to their food supply. At the four localities where they were found they were located in old grain fields and their burrows and habits appeared to Dipodomys phillipsi Gray. 80 be the same. In no instance was a group of holes noted, and it was rare to find two entrances to the same burrow. "Except at Tlalpam, they were not found in considerable numbers in the Valley of Mexico, and were scattered. Wherever found they were always in very loose, sandy soil. At Tlalpam their burrows appeared to be shared by the little yellow pocket mouse (Perognathus], which was nearly as common as the Dipodomys. "As usual in this group, these animals are strictly nocturnal, and judg- ing from the number of tracks, they must be very active during calm nights, even when the temperature is some degrees below freezing. For one or two nights during a severe storm they do not venture out, but if the 'storm continues longer they forage enough to procure food. At Ajusco at rare intervals a foot or two of snow covers their haunts, and cold storms are common. Elsewhere in their range sharp frosts are com- mon during the winter months. " The preceding notes concerning the habits of this species, although written with particular reference to the south end of the Valley of Mexico, apply equally well to other parts of its range. In the eastern part of its range it frequents the same open sandy fields, but many of the holes were also found at the bottom of shallow ditches entering the ground at the foot of the low bank at the sides. Others were found sim- ilarly situated along the sides of small arroyos. In one field above Chal- chicomula the holes entered the ground almost perpendicularly for five or six inches before sloping away at the usual slight angle. This was due, however, to the fact that the very loose sandy character of the sur- face soil made it impossible to start a hole in at the usual angle." Identity of Nelson's Specimens with Dipodomys phillipsi Gray. Apart from the geographic proximity of Real del Monte to the region where Mr. Nelson obtained his specimens, and wholly independent of the question as to whether or not the animals whose burrows were observed 6 miles south of Pachuca belong to the same species, or to the genus PerodipuSj the specimens themselves afford positive proof that they are the species de- scribed by Gray as Dipodomys phillipsi, as may be seen from Gray's original and very circumstantial description, which is here reproduced entire : "A new Genus of Mexican Glirinc Mammalia. Mr. John Phillips, who has latelv returned from Eeal del Monte, Mexico, has, at the recommen- dation of Mr. John Taylor, sent to the British Museum the skins of some very rare and interesting birds, of a Hassans, and of the new animal which I shall now proceed to describe. This animal is very interesting, as hav- ing all the external form and coloring of a Gerboa; and it is doubtless the American representative of that African genus, though differing from it very essentially in being provided, like some other American genera, 90 Merriam Rediscovery of the Mexican l\aii<> Hat us SaccophoruSj Saccomys, and //r// TO// ///.>, with large cheek-pouches, which open externally on the side of the cheeks. I propose to call it "Dipodomys. "Body covered with soft hair. Head moderate, with large cheek- pouches opening externally on the side of the cheeks. Ears and eyes rather large ; the fore-legs short ; the hind tarsus long and slender; the hind feet very long; the soles covered with hair; toes 5-4. The tail much longer than the body, covered with rather short hair, and with a dilated brush at the end ; the upper cutting teeth grooved in front. Grinders (?) "This genus differs from all those above cited in the tail being elongated and covered with hair, with a pencil at the ends like the Gerboas, and from Saccomys in the soles of the hind feet being hairy. "Dipodomys phiUipii Gray. " Gray-brown, with longer black hairs ; sides sandy ; sides of the nose, spot near the base of the ears, band across the thigh and beneath, pure white ; nose, spot at the base of the long black whiskers, and at the base of the tail, black ; tail black-brown, with the band on each of its sides and tip white; penis ending in a long spine- Length : body and head, 5 inches; tail, 6J inches; hind feet, !} inch. "Inhab. Mexico, near Real del Monte. John Phillips, Esq." (,T. E. Gray, Annals and Magazine of Natural History, vol. vn, No. 46, August, 1841, pp. 521-522.) By a fortunate circumstance Gray's type specimen was figured by Audubon in his colored plate (No. cxxx), which plate has the additional merit of tallying with the description and also with the specimens collected by Mr. Nelson.* The measurements given by Gray (and repeated by Audubon) are: Head and body, .5 inches ; tail, 61 inches; hind feet, U inches. Converted into millimeters, the length of the tail is 165, and of the hind foot 38. The averages of these measurements in 54 specimens collected by Mr. Nelson are : Tail, 167.5 ; hind foot, 41. In several individuals the hind foot fell to 39 mm. Mr. Nelson's measurements Avere taken in the flesh, which ac- counts for the slight difference in the length of the nihd foot. * Audubon states : " Our drawing was made from a beautiful specimen in the British Museum, which was the first one brought under the notice of naturalists, and the original of Mr. Gray's description of this singular animal ; it was procured near Real del Monte, in Mexico." (Quadrupeds of North America, vol. in, 1854, 140.) Dipodomys phillipsi Gray. \ 91 In the light of the ample material collected by Mr. Nelson, the species may be redefined as follows : Dipodomys phillipsi Gray. Dipodomys phillipii Gray, Ann. and Mag. Nat. Hist., vn, 1841, 522. (Name spelled phillipii by typographical error and corrected to D. phillipsii by same author a few months later. Am. Journ. Sci., XLII, 1842, 335.) Mncrocolus halticm A. Wagner, Wiegmann's Archiv fiir Naturgeschichte, 1846, 172-177 (from 'Mexico'). Jjljtodoini/s pltilUpsii Audubon and Bachman, Quadrupeds of North Amer- ica, vol. in, 1854, plate (colored) cxxx (of Gray's type speci- men), and pp. 137-140 in part. NOTE. The D. phillipsi of subsequent authors is composite and does not include this species at all. Type Locality. The Valley of Mexico, Mexico. General Characters. Tip of tail white; size, smallest of the species having this peculiarity (about equaling D. merriami, but tail longer) ; coloration very dark, resembling D. calif ornicus. Measurements. Average of 23 adult specimens from Tlalpam and Ajusco, at the south end of the Valley of Mexico (measured in the flesh) : Total length, 270 mm. ; tail vertebra, 168 ; hind foot, 41. Color. Upper parts sepia-brown, more or less suffused with ochraceous, and everywhere conspicuously mixed with black- tipped hairs ; thigh patches large, becoming dusky, and forming a large black patch behind and on the sides of lower leg and ankle, reaching upper surface of heel ;. crescents at base of whiskers large, nearly black, and meeting across nose ; eyelids black ; supraorbital spot obscured ; upper and lower tail stripes black, meeting along distal third, and succeeded by a short brush of pure white, usually measuring only 10 to 15 mm., thus being decidedly shorter than in any other white-tipped species. Cranial Characters. Skull slightly larger than D. merriami, but not so large as D. californicus. Viewed from above, the post-rostral portion of the cranium is subquadrate in shape, owing to the gi*eat breadth across the orbital bridges of the maxillaries, and the relatively slight development of the mastoids. The least interobital breadth of the frontals is greater than in any other species of the genus, equaling or exceeding the distance from the foramen magnum to the premolar, and considerably exceeding the length of the nasal bones. The tympanic capsule 13 BIOI-. Soc. WASH., Vor-. VIII, 1893. 02 Merrlam Rediscovery of the Mexican l\<(nO 1 I I ( I 1 T 1 1-3 - i I el O . 00 ifi O 1 H 1 8 . " S . $ -i O CO CO CO t^ CO 0> rH rH T 1 rH as- certain one or more of the following facts : (1) the normal* size of the animal; (^) the proportions (or ratios) of some of its parts, and (3) the limits or extremes of variation in a large series of individuals (which may be expressed both in actual maximum and minimum measurements and in percentages of the normal). The accompanying table not only summarizes this information for each locality in a single line, but also, by bringing into sharp contrast corresponding data from different localities, shows the amount of geographic variation in the species. Tables of the ordinary sort may contain the material from which these important facts can be ascertained, but the labor of digging them out from the bewildering mass of figures in which they are buried is so great that it is rarely undertaken. Note on Macrocolua halticus Wagner. A word is necessary, perhaps, concerning the Macrocolus halticus of Wagner,f which has been a stumbling block to naturalists for many years, though its true position was correctly indicated by Baird thirty-five years ago. Baird's remarks were as follows : "Although Wagner expressly states that there were no external cheek pouches in his specimen, and that in consequence it could not belong to Gray's genus, Dipodomys, yet the coincidence in every other respect skull, teeth, skeleton, and external form is so very intimate as to render it almost certain that the *The normal of any measurement is the average or mean of that mea- surement in a large series of adult individuals from the same locality. In preparing such tables care should be taken to exclude all immature and imperfect specimens. t Macrocolus halticus A. Wagner, Wiegmann's Archiv. fiir Naturgesch- ichte, 1846, 172-177. Dip odo mys phillipsi Gray. 95 cheek pouches must have been overlooked, especially as we are particularly informed that the specimen was in very defective condition as preserved in alcohol. The species was probably identical with that described by Gray, viz : D. phillipii, which appears to be the one common in Mexico." * It should be borne in mind, however, that at the time Baird wrote (in 1857) all of the kangaroo rats were referred to Dipodomys, the genus Perodipus (containing the 5-toed species) not having been established until ten years later.f That Wagner's animal could not have been a Perodipus is clear from the statement that it had only 4 toes on the hind foot, and by other details of the original description. On the other hand, that it was a true 4-toed Dipodomys is shown by Wagner's original description, which includes a detailed account, with measurements of the skull and skeleton ; J by his drawing of the skeleton (natural size), published a little later, and by his drawings of the teeth, contained in the supplement to Schreber's Saugthiere (tab. 239, E.). The genus being disposed of, the species now comes up for determination. The fact that Baird referred it to D. phillipsi of Gray, in which he was followed by Coues, is of no weight, since both of these authors included under this head not only several very distinct species 3 but also species belonging to both genera. Nothing is known of the source of the specimen save that it came from Mexico. Three species of the genus are now known to inhabit Mexico, namely, D. phillipsi, D. spectabilis, and D. mer- riami. D. spectabilis may be at once eliminated because of its much larger size, leaving only D. phillipsi and D. merriami be- tween which to choose. Wagner's measurements of the skull and skeleton give a total length of 269 mm. and a length of tail vertebraB of 168. His external measurements of the body and tail added together make a total of 279 mm., an excess of 10 mm. over the length of the skeleton, as usual when the body and tail are measured separately, the length of the body overlapping the point from which the tail measurement begins. Allowing s * Baird, Mammals of North America, 1857, 409. fFitxinger, Sitzungsber. math.-nat. classe, K. Akad. Wiss. Wien, LVI, 1867, 120. \ Macrocolus halticus A. Wagner, Wiegmann's Arcliiv. fur Naturgesch- ichte, 1846, 172-177. $ Beitrage zur Kenntniss der Siiugethiere Amerikas, II Abth., 1848, 319 und 332, pi. vn (Abhandl. d. II, Cl. d. k. Ak. Wiss V, Bd. II). 96 Merriam Rediscovery of the Mexican Kangaroo Eat. 1 mm. for the thickness of the skin over the end of the nose, the measurements given by Wagner for the skeleton of his Macrocolus halticus (converted into millimeters) are : Total length, 270 ; tail vertebrae, 168 ; hind foot, 39.5. The averages of 23 adult specimens of Dipodomys phillipsi collected in the Valley of Mexico by Mr. E. W. Nelson are: Total length, 270; tail vertebrae, 168 ; hind foot, 41.1, the agreement being surprising. The measurements and description of the skull and skeleton also agree with those of D. phillipsi and not with D. merriami or any other species. The number of tail vertebrae is stated by Wagner to be 31. In two skeletons from Tlalpam the number is 32. No other species of Dipodomys is known to have more than 29, and no species of Perodipus more than 28. Further- more, if additional evidence is needed, the drawing of the skull shows conclusively (if even approximately correct) that the animal is D. phillipsi, the interorbital breadth of the frontals being considerably greater than in any other species, and the mastoids being only slightly developed for a Dipodomys, the consequent breadth of the cranium across the mastoids but slightly exceeding the breadth across the maxillary bridges of the orbits. It has been shown by the process of elimination that Wagner's animal could not belong to any known species other than Dipodomys phillipsi; and it has been shown by direct compari- son of his description with duplicate types of the latter species that it does agree with this species in every particular. It seems safe to infer, therefore, that Macrocolus halticus of Wagner is Dipodomys phillipsi of Gray. VOL. VIII, PP. 97-108, PLS. XII, XIII JULY 20, 1893 PROCEEDINGS 4 OF THE BIOLOGICAL SOCIETY OF WASHINGTON. THE PALEONTOLOGY OF THE CRETACEOUS FORMA- TIONS OF TEXAS. THE INVERTEBRATE FOSSILS OF THE CAPRINA LIMESTONE BEDS. BY ROBERT T. HILL. CONTENTS. Page. I. Stratigraphic Position of the Caprina Limestone Beds in the Comanche Series 97 II. Characteristic Fossils 100 III. Age of the Caprina Limestone Beds 104 IV. Description of Species 105 V. Plates.. . 108 I. STRATIGRAPHIC POSITION OF THE CAPRINA LIMESTONE BEDS IN THE COMANCHE -SERIES. About midway in the column composing the Comanche series or Lower Cretaceous of Texas, and constituting the uppermost member of the Fredericksburg division (Comanche Peak group of Shumard in part), there is a peculiar group of strata known as the Caprina limestone of Shumard.* Dr. Shumard placed the bed in the Upper Cretaceous, at the very top of the whole of the fifteen or more subdivisions of the * First Annual Report Texas State Geological Survey, Austin, 1880-'88, pp. 124-126. 14-Bioi,. Soc. WASH., VOL. VIII, 1893. (97) 98 Hill The Invertebrate Fossils of t fie two great formations of Texas, instead of in the middle of the lower series, where it belongs. His original description of it is as follows : Caprina Limestone. This is the uppermost recognized member of the series and, although of no great thickness, has a somewhat extended geographical range. It is a yellowish-white limestone, sometimes of a finely granular texture and sometimes made up of rather coarse, subcrys- talline grains, cemented with a chalky paste. It generally occurs in thick massive beds, and is capable of withstanding the action of the weather to a greater extent than most of the members of the Cretaceous system. This formation is usually found capping the highest elevations, and its presence may be nearly always recognized, even at a distance, by the peculiar flat-topped and castellated appearance it imparts to the hills. According to Dr. Riddell, it is finely displayed along the bluffs of Brazos river in Bosque, McLennan, and Hill counties ; also along the Leon and Bosque rivers. The summits of the remarkable elevation known as Comanche peak, in Johnson county, and that of Shovel mountain, in Burnet county, consist of this rock. The fossils are chiefly Caprina, Cytherea, and Ammonites of undetermined species. (Trans. St. Louis Acad. Science, vol. i, 1860, pp. 583-584.) In a previous paper the writer has described the Caprina lime- stone substantially as follows : The Caprina Limestone. Without any serious stratigraphic break in the chalky limestones the abundant Comanche Peak fauna disappears and there continue 300 feet, more or less, of chalks and chalky lime- stones of varying degrees of consistency, from a pulverulent condition to firm limestones, which seem to be a secondary condition of the chalk produced by superficial hardening. These hard layers form the table rock of the buttes and mesas of the extensive Grand Prairie region and are exposed in the river bluffs between Austin and Mount Bonnell, on the Colorado, where the chalk has been more or less hardened into firm limestones by the local metamorphisrn accompanying faulting. The lime-kilns and quarries immediately west of Austin are all located on this subdivision. Accompanying these chalks and chalky limestones are well defined layers of exquisite flint nodules, occupying apparently persistent horizons in localities. The flint nodules are oval and kidney-shaped, ranging in size from that of a walnut to about two feet in diameter. Exteriorly they are chalky white, resembling in general character the flint nodules of the English chalk cliffs. Interiorly they are of various shades of color, from light opalescent to black, sometimes showing a banded struc- ture. These flint nodules are beautifully displayed in situ in the Deep Eddy canyon of the Colorado, above Austin, where they can be seen occupying three distinct belts in the white chalky limestones. Where these chalky limestones form the basis of extensive plateaus, such as the remnants of the Grand Prairie west and southwest of Austin, Caprina Limestone Beds. ( .> ( .) the flints are left in great quantities as a residuum of the softer chalks (which are more readily decomposed into soils and washed away), and they cover large areas of country. They have also been transported eastward in past geologic times by marine and river action, and are dis- tributed over large areas along the margin of the Black Prairie region as a part of the Post-Cretaceous gravels of that region. In some of these flints remarkable decomposition is exhibited, the products being geode- like cavities lined with quartz-crystals and pulverulent substance. These are the only flint horizons, so far, at least, as is known to the writer, in the whole of the immense Cretaceous deposits of the United States. They occur about the middle of the r Lower Cretaceous series instead of at the top of the Upper series, as in England. It was from these flints that the ancient and modern Indians made their flint imple- ments, and the ease of their lithologic identity will be of value to the anthropologist in tracing the extent of the intercourse and depredations of former Indian tribes inhabiting this region. Occasionally the flints, especially an opalescent variety in Comanche county, possess nuclei in the shape of fossils, usually Requienia. The decomposition of these flints and of the adjacent limestones has produced some peculiar and unique effects in the rocks and landscape of the region, the silica replacing the calcium carbonate and leaving as a remnant a peculiar porous cavernous rock, usually of a deep-red color from the hydration of the iron pyrites into limonite, composed of the siliceous pseudomorphs of fossil Rudistes and other shells, the interstitial spaces glittering with minute quartz crystals which line them. Immediately west of Austin, along the downthrow of the great Bonnell fault in the bluffs of the Colorado, another peculiar transformation takes place in the Caprina limestone. Occasional red decomposing spots occur in the massive white chalky limestones. Upon closer examination the apparently non-fossiliferous limestone is seen to be undergoing decompo- sition into a dry pulverulent inflorescence, and as a residuum there re- mains a dry red dust containing exquisitely preserved calcite pseudo- morphs of many rare fossil?, such as recently described by Roemer and White, the occurrence of which I have located in this horizon. Traces of the following economic products have been discovered : Pot- ash, salt, strontianite, anhydrite, epsom salts, gypsum, and gold, but in quantities as yet unknown. At Austin a fault of about 750 feet downthrow has broken this lime- stone division into two different areas, and hitherto confused its measure- ment. The limestones are more resistant to erosion than the over and under lying strata, and hence form the summit scarps and mesas of the peculiar buttes and divides in Hood, Comanche, Hamilton, Bosque, Coryell, Lampasas, and other counties of the Grand Prairie regions of Central Texas. It also occurs as the surface of extensive prairie regions in western Williamson county. It also caps the summit of thS Jehosaphat plateau in 100 Hill The Invertebrate Fossils of the northwestern Travis county, and the Edwards plateau to the south, where its surface outcrops, owing to rain sculpture, is weathered into extensive fields of " Karrenfelder " or miniature mountains. The limitations of this group of strata have not been finally determined, but it should include as its upper members the Austin marble (the Upper Caprotina limestone and the litho- graphic flags of my local Austin section). No abrupt break is evident between these and the underlying beds which contain the Comanche Peak fauna of Shumard (Die Kriedebildungen bei Fredericksburg of Roemer). The detail of these beds at Austin have been given by Mr. J. A. Taff (who made the section under my supervision) in the Third Annual Report of the Texas Geological Survey. Stratigraphically the Caprina limestone represents the culmi- nation of the subsidence that progressed during the Comanche epoch. Paleontologically the Caprina limestone beds are of the great- est interest, for in them we have the development of the aber- rant Chamidas and Rudistes of this country. They contain all the species of these families known to occur in the Cretaceous of the United States, with the two exceptions of the Caprotina-like Coralliochama of ^California and the large Radiolites austinensis- like forms so common in the equivalents of the Colorado group of the Upper Cretaceous in the Alabama, Texas, and Colorado regions. As it is clearly and distinctly overlain by the whole of the Washita division which corresponds to the Gault of Europe, as will be later shown by the writer, and is above the well- defined beds of the Trinity and Lower Fredericksburg division, it affords an important landmark in tracing the progress of marine life on this side of the ocean. II. CHARACTERISTIC FOSSILS. The fauna of the Caprina limestone has been little understood, owing to the unfortunate fact that many of its fossils have been attributed to other horizons. Shumard* included the Lower Caprotina limestone of the Trinity division in Hood county with the Caprotina limestone of the Caprina beds, and through- out his valuable literature one fails to find any distinction be- tween them. A few years since my friend, Mr. George Stolley, * Loc. cit., p. 584. Caprina Limestone Beds. 101 discovered a fauna in what is now known to be the Caprina limestone in the bluffs of the Colorado, a few miles west of Austin. Not knowing the horizon of these beds, he sent them to Dr. C. A. White, of the United States Geological Survey, and Dr. Ferdinand Roemer, at Breslau. Dr. White described sev- eral of the forms of aberrant Chamidge, but owing to the lack of stratigraphic particulars he refrained from publishing any age conclusions.* Dr. Roemer published f beautiful illustrations of these fossils and described them with his accustomed skill, but at that time, not knowing the comprehensive character of the beds in the immediate vicinity of Austin and probably deceived by the lithologic resemblance of the matrix, he erroneously con- cluded that they came from the Austin chalk (Niobrara) of the Upper Cretaceous. His conclusions that the Austin chalk from which he supposed these fossils to have come was of Turonien age was undoubtedly correct. The Austin chalk abounds in many other species which justify such a conclusion, but not one of these species later described warranted such a conclusion, nor did they come from that horizon. The writer regrets that he is unable here to republish Dr. Roemers excellent figures and descriptions of this fauna. The following is a list of the species which I have observed from the Rudistes horizon in the Caprina limestone beds at Aus- tin, Texas : Parasmilia austinensis Roemer. Pleurocora texana Roemer. Pleurocora coalescens Roemer. Cladophyllia furcifera Roemer. Coelosmilia americana Roemer. Holectypus Roemer. Patella or Pileolus (?). Chrysostoma. Helicocryptm or Adeorbis. Ziziphii ins ( Calliosto-ma) . Nerinca austinensis Roemer. Nerinea cidtrispira Roemer. * Bulletin 4, TJ. S. Geological Survey, on Mesozoic fossils, by C. A. White, .Washington, 1884. t Paleontologische Abhandlungen Herausgegeben von W. Dames und E. Kayser, Vierter Band. Heft 4. Ueber eine Durch die Haeufigkeit Hippuriten-Artiger Chamiden Ausgezeichnete Fauna der Oberturonen Kreide von Texas von Ferdinand Roemer. Mit 3 Tafeln. Berlin, 1888. 102 Hill The Invertebrate Fossils of the Nerinea subula Roenier. Glaucoma (?). Cerithium obliterato-granosum Roenier. Cerithium austinensis Roemer. Trochus texanus Roemer. Solarium planorbis Roemer. Natica (Amauropsis) avellana Roemer. Requienia patagiaia Ch. A. White. Monopleura marcida Ch. A. White. Monopleura pingaiscida Ch. A. White. Lucina acute-lineolata.RoQmer. Requienia patagiata Ch. A. White. The following forms have a more general occurrence : Oslrea munsoni sp. nov. Radiolites texana Roemer. " dai'idsoni sp. nov. Requienia texana Roemer. Ichthyosarcolithes angais Roemer. Monopleura marcida Ch. A. White. Ammonites (Buchiceras) pedernalis von Buch. (Schloenbachia) acute-carniatiis von Buch. In addition to the foregoing numerous species have been described under the generic name of Caprina, owing to the occur- rence in immense quantities of a fossil supposed to have belonged to that genus. These fossils, however, are usually imperfectly preserved, but it can now be said with assurance that none of them belong to that genus, but are mostly Ichthyosarcolithes or Radiolites. All of the so-called Caprinas heretofore described from Texas come from this horizon. Most of the Austin species occur in the bluffs of the south bank of the Colorado and Barton creek, just west of Austin, as beautifully preserved calcite pseud omorphs. Usually the lime- stone is very barren of all fossils except the Rudistes and Chamidse. Other aberrant Chamidse and Rudistes from the Texas Creta- ceous have long been known, but their exact stratigraphic range has not been clearly stated. With the exception of Radiolites davidsoni herein described, the stratigraphic occurrence of all the species was unknown to their authors when they described them. Many were described from imperfect . specimens, and all the writers previous to Dr. White's valuable contribution expressed Caprina Limestone Beds. 103 serious doubts as to the true generic position of the forms. It can now be said that, with the single exception of Radiolites austinensis, all of these forms in Texas come from the Caprina limestone. The following is a list of the forms thus far described : CHAMID^E. Dicer as (?) Roemer. Requienia bicornis Meek, 1876. Fort Lancaster, Texas. patagiata White, 1884. Near Austin, Texas. texana Roem., 1852 ; White, 1884. Near Austin, Texas. Highlands between New Braunfels and Fredericks- burg. Marcou, 1858, reports this form at " Comet creek, on left bank of the False Washita." Monopleura marcida White, 1884. Near Austin, Texas. " subtriquetra Roem., 1852. Valley of San Saba and upper arm of Pedernales river. pingiuscula White, 1884. Near Austin, Texas. " texana Roem., 1852. Ichthyosarcolithes anguis Roem., 1888, Barton creek, west of Austin. (?) (Caprina) crassifibra Roem., 1849, 1852. (?) (Caprina) guadalupx Roem., 1849, 1852. (?) (Caprina) planata Con., 1855. Oak creek, near Pecos, Texas. " (?) (Caprina) occidental Con., 1855, 1857. Pecos river near mouth. (A. Schott.) (?) (Caprina) texana Roem., 1849. Plagioptychus (?) cordatus Roem., 1888. Radiolites (Hippurites) texanus Roem., 1849, 1852. davidsoni sp. nov. All of the above species occur in the Caprina limestone. Radiolites austinensis Roemer, is the only other form from the Texas Cretaceous. It occurs in the Austin chalk, and is so radically different in every aspect that it hardly belongs in the same group with the lower forms. With the exception of Mo- nopleura and Requienia, which range downward into the Trinity Division, all the other genera occur only in the Caprina lime- stone, appearing suddenly upon the scene with these beds and completely vanishing thereafter. Radiolites texanus Roemer, which comes from the Caprina 104 Hill The Invertebrate Fossils of tlie limestone, was referred to the genus Hippurites by Roomer,* although upon reading his description as originally published it will be seen that he distinctly stated that it was exceedingly doubtful whether this form belonged to llddta/itc* or Hippurites. The name Hippurites, however, has gone forth in literature, and, inasmuch as this genus is a characteristic form of the Upper Cretaceous of Europe, its supposed occurrence in the Lower Cretaceous of Texas has been the greatest obstacle to man in accepting the lower position of the Comanche Series. I am now prepared to state that there is not a single Hippurites '\ in either the Lower or Upper Cretaceous of Texas, and that this unfortu- nate impression should no longer prevail. III. AGE OF THE CAPRINA LIMESTONE. The writer does not feel prepared to separate the Caprina limestone from the remainder of the Fredericksburg Division as a unit for the discussion of homotaxy, and in the following remarks it should be remembered that the beds are stratigrapli- ically related. Dr. Fred. Roemer, in his classical monograph of the Kreidebil- dungen von Texas, placed the beds which are now known to be the Caprina limestone at the very top of the Texas Cretaceous and referred them to the Senonien. Forty years later he un- knowingly described more of the fauna from the same beds, and placed them in the Turonien.J In earlier writings I have shown the erroneous impression under which Dr. Roemer thus placed these beds, and that instead of occurring above his Tu- ronien (Austin chalk beds) they are stratigraphically below them, and hence could not be Senonien. Shumard, who first defined and applied the present name to the beds whose fossils had been described in part by Roemer, also failed to discover the true stratigraphic position, and like- wise placed them at the top of the Texas Cretaceous. || * Kreide, von Texas, p. 76. fin the Third Annual Report of the Texas Geological Survey a species is mentioned by name only as "Hippurites flabdlata sp. nov." from the Caprina limestone. No description whatever has been given of this form. From the writer's familiarity with the specimens in the Texas collection he thinks it probable that it must be a Radiolites. $ Paleontologische abhandlnngen. % Am. Journ. Sci., vol. xxxvir, 1880, pp. 318-319. Ibid, April, ISJK'J. || Loc. cit. Caprina Limestone Bed*. 105 The writer has repeatedly shown that the stratigraphic posi- tion of the beds was in the middle of the Lower Cretaceous or Comanche series instead of at the top of the Upper, as believed by Roemer and Shumard, and hence, aside from the paleon- tologic evidence, he would assign these beds to a still lower horizon, probably the Uppermost Neocomian, or Transitional Neocomian-Gault, for the following reasons : 1. The fauna does not contain a single characteristic genus or species of beds of higher position. 2. The beds occur immediately beneath the Washita Divis- ion, which contain numerous species resembling those of the Gault of Europe. 3. The beds bear a remarkable paleontologic and stratigraphic resemblance to the Requienia Limestone beds of France and the Spanish Peninsula, where similar limestones, with Radiolites&nd Requienia. abound in the Upper Neocomian. IV. DESCRIPTION OF SPECIES. Ostrea munsoni sp. nov. Plate XII. Compare 0. Joanx Choffat. Eecueil de Monographies Stratigraphiques Sur le Systems Cretacique du Portugal, par Paul Choffat. Lisbon, 1885, p. 34, plate 1, figs. 1-7. Very thin and flat; elongately sub-triangulate and marked by many well defined radiating ribs; the pallial extremity rounded ; beak more or less acuminate and slightly deflected, and evidently slightly attached ; the inferior valve slightly con- cave, nearly flat, and showing near its beak an area of attach- ment. The larger valve flatly convexed and only slightly larger than the lower; the ornamentation of both valves is sim- ilar, and as remarked by Choffat in his description of Ostrea Joanne, a very similar form from Portugal, the two valves present an appearance as if they had been plicated together, the one upon the other. Each shell is very thin, and the living space small. When closed together the thickness of both valves is hardly one-twentieth the length of the shell. The finely fluted ribs are slightly sinuous, continuous from beak to base or sometimes bifurcated, alternating with short ribs extending only halfway from base. This is especially true upon the lower valve. This species is easily distinguishable from all the other North American oysters by its extreme flatness and thinness. 15 BIOL. Soc. WASH., VOL. VIII, 1893. 106 Hill The Invertebrate Fossils of the I have observed only a few specimens. These occur with great scarcity in the Caprina limestone, in association with Radiolites davidsoni at Belton and near Austin. The species is named in honor of Mr. T. V. Munson, of Denison, Texas. This species resembles in general form the excellent figures of 0. Joanae Choffat, from the Cretaceous of Portugal. Radiolites davidsoni $p. nov. Plate XIII. Description of Figured Specimen. Very elongate, measuring over 40 centimeters ; slightly flaring at larger end and gently tapering to small extremity ; somewhat sinuous, pointed, and attached at lower end ; exterior surface strongly marked by longitudinal ribs and grooves, as follows : Two especially broad and conspic- uous grooves extending the entire length of the shell on oppo- site side to that shown in large figure, but seen in small segment and end view. These are accompanied by two large corre- sponding oblate ribs and a sharp, angular carina. These two grooves, ribs, and carina occupy one of the sides of the sub- triangular circumference. The carina is very prominent and sharp, and extends from the smaller end to within about ten centimeters of the large extremity, where it becomes exfoliated and deflected like a mantle over and across the large ribs and grooves. This carina forms the upper margin of the large figure, and the angles seen in the cross-sections, and is opposite the smaller angle of the semi-lunate interior. It is bordered upon one side by one of the major grooves above described and upon the other, by the somewhat flattened face shown in the figure of larger cross-section, on Plate XIII. The remaining two-thirds of the surface of the shell is marked by about fifteen small linear ribs, separated by wide, slightly concave depres- sions. Interior of shell a hollow cavit}'' subpyriform or semilu- nate in cross-section, and marked by a few widely separated concave partitions occurring at intervals of three to five centi- meters. Opposite the left hand major groove there is a slight projection extending the length of the interior which makes .a narrow sinus in the casts when found. The interior is usually filled with calcite crystals. Shell, irregularly thick, varying from three-quarters to one- half centimeter in different parts "of the circumference. Outer surface is very smooth between the flutings and marked by fine Caprina Limestone Beds. 107 cross-stria3, which are the exterior terminals of the circum- scribant septa? of the honeycombed structure. The shell is composed of two layers (see fig. 1). The outer one (a) is very thick and constitutes most of the substance and has a fine cellu- lar honeycombed structure. Cells very minute and rectangular in cross-section and produced by the intersection of the concen- tric and vertical lamina?. The imbricate concentric lamina? are arranged in successive layers diverging upward from the interior layer of the shell. The interior shell (6) is thin and very poorly preserved, being largely replaced by calcite crystals. FIG. 1. Longitudinal section of larger extremity of specimen figured on plate .I'iii. a, outer shell of larger valve ; b, space once occupied by inner shell of same; c, a cross-septum of interior cavity; d, last chamber of interior cavity ; e/ e", calcined area, marking position of the muscular apophysis of upper valve ; /, small fragment of upper valve ; /', section of outer edge of apophysis ; g, undetermined fragment. The opening of the shell is composed of the thin interior shell and a few layers of the exterior shell. Most of the con- centric lamina? of the latter gradually disappear before reaching this termination. The dwelling chamber is about one and one- half centimeters deep, and the details of its structure somewhat concealed by the filled-in matrix. A longitudinal section of the large extremity (fig. 1) gives no detail of the anatomy of the living chamber, but gives some light on the upper valve. 108 Hill Fossils of the Caprina Limestone Bads. Upper Valve. No satisfactory specimen of the smaller valve has been found. In the specimen figured on plate XIII and in accompanying figure at / there is a fragment preserved, super- ficially resembling the shell of a Pecten, which, however, upon careful inspection appears to be a broken remnant of the smaller upper valve. It is composed of strongly radiating ribs alternating with finer lines, and in the longitudinal section shows a well-defined apophysis (/) corresponding to one of the casts preserved in the larger valve (e"). The section also shows distinctly the casts of the two muscular apophyses (e e") of the small valve, as in Radiolites. The generic position of this form has been perplexing, and possibly it deserves a distinct generic position. That it is nearer Radiolites than Hippurites is clearly shown by the absence of many of the characteristic features of the latter genus, such as the numerous partitions which cross the interior cavity ; the different structure of the dwelling-chamber, and the presence of only two instead of three longitudinal sutures. Upon the other hand, it possesses many of the distinguishing characteristics of Radiolites, such as the prominence of the two well-defined lon- gitudinal sutures and the structure of the interior cavity. It also differs from the genus Sphserulites, which is characterized by one longitudinal suture. The cells are mostly rectangular, while of the'genus Radiolites, according to Zittel, there are five or more sided. This distinction has been used by some writers to make new genera, but the writer prefers to defer such action. The form occurs in great abundance in the limestones near the water's edge of the Lampasas and Leon creeks, in the eastern suburbs of Belton, Texas, from whence the type specimens were collected by Professor Wilson T. Davidson, in whose honor it is named. It is also abundant near Round Rock and Austin. There is a possibility that this species may be the same as the form entitled "Hippuritet texanus" of Roemer, figured and de- scribed in Die Kriedebildungen von Texas, of Roemer ; but how- ever strong the inclination may be to think they are the same, his descriptions and figures are so radically different that they cannot be said to be identical. H. texanus, as figured, shows the cells to be polyginal instead of rhomboidal, as in R. dachhoni, and the surface grooving and cross-sections are entirely different in detail. R. davidsoni shows great variation in length, some of the forms being very stunted and much thicker than the specimen figured. PLATE XII. Ostrea munsoni sp. nov. Three figures showing one view of an adult form and two views of medium-sized specimen. The older specimen shows larger valve on top exposing portion of lower valve at pallial margin. Both sides of the smaller specimen are shown, the smaller valve at the upper right-hand corner of the plate and the larger valve at the lower left-hand corner. Puoc. BIOL. Sue. WASH., VOL. VIII, 189.,. PLATE XII. OSTEEA MUNSONI Sp. 11OV. PLATE XIII. Radiolites davidsoni sp. nov. The illustrations^how an elongated specimen with details of its general structure. In the rower right-hand corner is a figure showing the two main sutures, which extend the entire length of the shell, but are not seen in the large illustration. Jnthe lower left-hand corner is an end view of the specimen figured. The object in the upper right-hand por- tion of this figure is a fragment of the broken upper valve. PRO( VOL. VIII, pp. 109-112 JULY 31, 1893 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON. TWO NEW WOOD RATS FROM THE PLATEAU REGION OF ARIZONA (NEOTOMA PINETORUM AND N. ARIZONA) WITH REMARKS ON THE VALIDITY OF THE GENUS TEONOMA OF GRAY. BY C. HART MERRIAM, M. I). The two new mammals herein described are of unusual inter- est, not only because they are very distinct from any heretofore recognized, but also because they inhabit a region that has been pretty thoroughly explored during the past few years, and from which an unusually large number of mammals have been already described.* One of the new species (N. arizonse) presents a remarkable combination of the external characters of the bushy-tailed wood rats (genus Teonoma of Gray) with the cranial characters of the round-tailed species (Neotoma proper). The other (N.pineloruni), is a round-tailed species allied to the N. fuscipes group of Cali- fornia. Incidentally, the study of A 7 ", arizonse led to the discovery of an important character that serves to distinguish Teonoma from Neotoma. * No less than twenty new species and subspecies were discovered in my biological survey of the San Francisco Mountain region in Arizona in 1889 (see N. Am. Fauna, No. 3. Sept., 189'J) ; and others have been de- scribed from the same general region by Dr. Edgar A. Mearns and Dr. J. A. Allen. 1C BIOT.. Soc. WASH., VOL. VIII, 1893. (109) 110 Merriam Two New Wood Rats from the Neotoma arizonee sp. nov. Type from KEAMS CANON, APACHE COUNTY, ARIZONA. No. J|| 9 ad. Merriam collection. Collected by J. Sullivan May 21, 1888. General Characters. Tail bushy as in Neotoma cinerea, but nar- rower ; animal similar to N. cinerea in general appearance, but smaller, and agreeing with the round-tailed species in important cranial characters. Ears large, measuring about 34 mm. from anterior base ; whiskers long and coarse, reaching shoulders. Total length about 365 mm. ; hind foot about 35 mm. (in an old male from the same locality the hind foot measures 39 mm.). Except for its superficial resemblance to N. cinerea, this animal needs no comparison with any known species. Color. Upper parts everywhere bright ochraceous-buff, mod- erately mixed with black-tipped hairs ; under parts and feet pure white. Tail bicolor : grayish brown above, white below. Cranial and Dental Characters. Skull smaller and much shorter than that of N. cinerea, with shorter nasals and nasal branches of premaxillaries, and larger and much more inflated audital bulke. The most important cranial character, however, con- trasted with N. cinerea, is the presence of a broad slit-like open- ing on each side of the presphenoid and anterior third of the basisphenoid, as in the round-tailed species generally. In the bushy-tailed N. cinerea this slit is completely closed by the as- cending wings of the palatine bones. The molar teeth are actually as large as and relatively larger than in N. cinerea. The enamel pattern of the last upper molar is a nearly perfect trefoil, though the posterior reentrant angle on the outer side is shallower than the anterior. Variation. The males are considerably larger than the females, and the young are gray in color, as usual in the genus. The Reams Canon specimens are very uniform in color. An imma- ture male from Tres Piedras, Taos County, New Mexico, collected by J. Alden Loring July 4, 1892 (No. 53016, United States Na- tional Museum, Department of Agriculture collection), has the upper parts gray, tinged with buffy-ochraceous, and the white of the under parts clouded posteriorly from the plumbeous of the under fur. A specimen from Fort Wingate, New Mexico, collected by Dr. R.W. Shufeldt (No. 3358 $ , Merriam collection) is a little older, but evidently not adult. It has the upper parts more strongly suffused with pale ochraceous and the belly white. Its tail is less bushy than any of the other specimens examined. A young-adult female, collected at Winslow, Arizona, by Clark Plateau Region of Arizona. Ill P. Streator (No. 53517, United States National Museum, Depart- ment of Agriculture collection) has the under parts stained a deep salmon pink from the soil. This color washes out, leaving the belly pure white as in the Keams Canon specimens. Geographic Distribution and Faunal Position. Neotoma anzonse inhabits the Tusayan or Moki district in eastern Arizona, and adjacent parts of the Painted Desert on the west, and New Mexico on the east. In all, eleven specimens have been examined: eight from the Moki country, one from Winslow in the Painted Desert, one from Tres Piedras in northern New Mexico, and one from Fort Wingate, near the western border of New Mexico. The species evidently belongs to the Upper Sonoran zone. Neotoma pinetorum sp. nov. Type from SAN FRANCISCO MOUNTAIN, A KIZONA . No. \ 1 1 \ f 9 ad. United States National Museum, Department of Agriculture collection. Collected by Vernon Bailey August 1(5, 1889 (original number, 866). Measurements of Type Specimen (taken in flesh). Total length, 355; tail vertebrae, 163 ; hind foot, 37 ; ear from anterior base, 25 (measured in dry skin). Average measurements of four adult specimens total length, 362 ; tail vertebse, 166; hind foot, 36.7. General Characters. Similar to N.fuscipes from southern Cali- fornia, but averaging slightly smaller ; tail, ears, and hind feet shorter ; tail more hairy ; hind feet pure white [not clouded with dusky as in N. fuscipes] ; back more strongly suffused with ful- vous ; whiskers long, reaching shoulders. Color. Upper parts fulvous, strongly mixed with black-tipped hairs ; face from nose to above eyes gray, slightly grizzled, color of upper parts stopping at (or above) wrists and at ankles ; under parts and feet white ; tail sharply bicolor, above blackish, below white. Other specimens are less strongly fulvous, and the young are gray. Cranial and Dental Characters. Skull similar to that of the southern form of N. fuscipes but broader; brain case broader and shorter ; palate and inter-pterygoid fossa longer ; teeth larger (both molars and incisors). General Remarks. In my report on the mammals of San Fran- cisco Mountain, Arizona, I referred the wood rat of the region to Neotoma mexicana of Baird.* At that time no typical specimens of mexicana were available for comparison. Since then, however? 14 specimens have been secured from the type locality, Chi- huahua city, Mexico. * North Am. Fauna, No. 3, Sept., 1890, p. 67, pi. x, fifL 5-8 (skull). 112 Merriam Two New Wood Rats. On reexamining the specimens from San Francisco Mountain and comparing them with the duplicate types of N. mexicana they are found to be widely dissimilar, not only in external appearance, but also in cranial and dental characters. Singularly enough, the new species resembles N.fuscipes of California much more closely than mexicana. It is an inhabitant of the pine belt at the base of the mountain, and therefore belongs to the Transi- tion zone. Its range probably meets that of N. mexicana, which inhabits the adjacent Painted Desert on the east and the Grand Canon of the Colorado on the north. The avenue by which its ancestors originally reached the pine plateau region of Arizona becomes a geographic problem of no little interest. Its nearest ally, the southern form of N. fuscipes, is closely restricted to California, and no member of thefuscipes group has- ever been reported from the Rocky Mountain region. Remarks on the Validity of the Genus Teonoma of Gray. In 1843 J. E. Gray separated the bushy-tailed wood rats gener- ically from the round-tailed species under the name Teonoma* In this arrangement he was followed by Fitzinger in 1867f, but not by other writers. The only character ever assigned the genus Teonoma, so far as I have been able to ascertain, is its bushy -tail, in contradistinction to the short-haired terete tail of Neotoma proper. I take pleasure, therefore, in calling attention to an im- portant cranial character which seems to have been overlooked. In the skulls of the round-tailed wood rats there is a long open slit on each side of the presphenoid and anterior third of the basisphenoid. These openings may be designated the spheno- palatine vacuities. In the bushy-tailed species {Neotoma cinerea and N. cinerea occidental-is) these vacuities are absent, being com- pletely closed by the ascending wings of the palatine bones. Whether this character is of generic or subgeneric value may be left an, open question for the present ; that it is a character of considerable importance cannot be denied. The most interesting feature connected with the new Neotoma arizonse herein described is that while it agrees with Teonoma in the possession of a bushy tail, it has the spheno-palatine vacui- ties of Neotoma proper. *List of Specimens of Mammalia in the British Museum, 184.', p. 117. t Sitzungsberichte Math. -Nat., Cl. K. Akad. Wiss.Wien, vol. LVI, Abth. i, 1867, p. 77. VOL. VIII, PP. 113-116 AUGUST 16, 1893 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON. NOTES ON THOMOMYS BULBIVORUS. BY GERRIT S. MILLER, JR. Thomomys bulbivorus was accurately described by Richardson more than sixty years ago,* but has remained, save for this author's account, entirely unknown to naturalists almost to the present day. Richardson's description of Diplostoma (?) bulbi- vorum was based on a " Camas-rat " from the " banks of the Columbia," a region of which the mammalian fauna has until recently been very imperfectly known; hence it is no surprising circumstance that this gopher has been so long overlooked by collectors. In the spring of 1890 Mr. A. W. Anthony took, at Beaverton, Washington county, Oregon, three specimens of Thomomys that soon after came into my hands and were immediately identi- fied with the subject of Richardson's description. Lack of proper material to determine the questions of nomenclature raised by this discovery prevented any publication at the time, and the matter was allowed to rest.' Recently, however, Dr. J. A. Allen, with an abundant supply of specimens at his command, re- viewed a number of vexed questions concerning the synon- omy of various species of Thomomys, and at his request the Anthony skins were placed at his disposal, the rediscovery of this long-lost animal being soon after announced,t while the * Fauna Boreali- Americana, I, 1829, 206. fBull. Am. Mus. Nat. Hist., v, 56, author's edition, published April 28, 1893. 17-BioL. Soc. WASH., VOL. VIII, 1893. (113) 114 Miller Notes on Thomomys bulbivorus. Thomomys bulbivorus of Baird and subsequent authors was referred to the Oryctomys (Saccophonis} bottx of Eydoux and Gervais. Although the rediscovery of this fine gopher has thus heen made known, it still remains to redescribe the animal, which I propose to do as follows : Thomomys bulbivorus (Richardson). Diplostoma (?) bnlbivorum Richardson. Fauna Bor.-Am., i, 1829, 206, pi. xviiiB (lettered "douglasii" by mistake). Thomomys bulbivorus Allen. Bull. Am. Mus. Nat. Hist., v, April 28, 1893, 56, pi. i, fig. 14 (skull). Specific Characters. Largest known species of Thomomys; colors very dark ; white markings about mouth extensive and in striking contrast ; tail almost naked ; skull exceedingly large and heavy. Adult ( 9 No. || J, collection of Gerrit S. Miller, Jr., Beaverton, Washington county, Oregon, May 5, 1890; A.W. Anthony, collec- tor) ; dorsal surface mixed clove-brown and yellowish chestnut, the hairs everywhere slaty plumbeous at base, the three colors indescribably blended, but the clove-brown predominating on the head and mid-dorsal region though without forming a distinct dorsal stripe, this giving way on the sides to the chestnut, which in turn is replaced on the belly by slaty plumbeous ; narrow ring around ear, muzzle, lips, outer edge of cheek-pouches, and ill-defined area extending thence to front legs very dark brown, almost black ; linings of cheek-pouches and broad space between white, in striking contrast with surrounding color ; a small white anal spot; dorsum of manus brownish, of pes white; a white tuft at proximal base of large tubercle on palm ; tail very sparsely clothed on basal third with brownish hairs, which are not sufficiently numerous to conceal the skin ; this in the dried specimen yellowish white, dark brown for 10 mm. at tip. The three specimens differ but little among themselves in color, the variation, such as it is, being due to the varying amount of clove-brown in the fur of the back. This is a trifle less in the two males than in the female. In one of the males (No. jj|?) there is an indistinct wash of mars-brown on the belly. The dorsum of. the right hind foot of No. \\\ is covered with brown hairs. This, however, must be purely accidental. Unfortunately the specimens were not measured in the flesh, but as they have been prepared with much care the following measurements taken from the dry skins are not without value : Notes on Thomomys bulbivorus. 115 No. Sex. Date. Length. Tail. Hind foot. HI 9 May 5, 1890 260 74 37 HI $ May 12, 1890 280 73 41 11? c? May 12, 1890 255 67 38 The longest hind foot among thirteen specimens of T. bottse from Nicasio, Marin county, California, is 32 mm. ; shortest, 28 mm. ; average. 29.7 mm. Thomomys bulbivorus differs from T. bottse so greatly in color, as well as in size, that a detailed comparison of the two animals is scarcely necessary. In T. bottse the prevailing tint throughout is wood-brawn, more or less mixed with russet dorsally and blackening about the mouth, muzzle, and cheek-pouches. The latter are here, as in T. bulbivorus, lined with white ; the area between, however, is usually dusky, sometimes more or less marked with white, but never, or at least very exceptionally, wholly white. The skull of Thomomys bulbivorus, in addition to its very much greater size, differs from that of T. bottte in many details of structure. The occipital portion is broader and flatter (ratio of height from inferior lip of foramen magum to mastoid width 50 in bulbivorus, 54 in bottse} and the fronto-palatal depth propor- tionally greater. The dorsal aspect shows no decided points of difference, though in T. bulbivorus all ridges and muscular at- tachments are more strongly accentuated. Ventrally, however, important differences at once present themselves. That surface of the exoccipital which appears on the ventral aspect of the skull immediately laterad of the condyle is in T. bulbivorus oc- cupied by a deep groove running obliquely to the axis of the skull, while in T. bottse the surface is almost -flat. The basi- occipital is much broader in proportion to its length in bulbivorus than in bottse; the audital bullse of the former are much flatter and less inflated than in the latter. The form of the pterygoids differs markedly in the two species, those of T. bulbivorus being much the larger and strongly concave internally with hamulars con- verging at the tips, while in T. bottse these bones are flat, with hamulars divergent posteriorly. Both foramen magnum and external nares are broader in proportion to their height in T. bulbivorus than in T. bottse. Except in size, the mandibles and teeth of the two species show no distinctive characters. The following table of cranial measurements and ratios of Thomomys bulbivorus and T. bottse will serve to illustrate some of these differences in greater detail. 116 Miller Notes un Thomomys bnlbivorux. Cranial Measurements and Ratios of T. buJhirortiH <>tt;r. T. bulbivorus. 416 3 T. f> 1183 c? tffe 1184 (? 1187 c? Number 225 9 226 tf 227 c? Sex Basilar length 53 50 39 49 46.2 51 48 35.5 30 7.4 20 22.2 35.4 15.4 22.4 6.4 7.4 11 10.6 9.8 11.4 40 9.6 11 73.95 62.50 46.60 34.41 41.60 20.41 87.56 103.77 44 41.6 30.6 24 6.8 16.8 18 29 13 17 6 5.2 10.4 8.2 8.4 9.2 33.6 8.8 10.2 73.55 57.69 40.86 31.25 37.98 21.92 115.38 126.82 40.6 38.2 29 21.4 6 15.8 16.6 27 12.2 16.4 5.8 5.4 9.8 8 8 9 30.2 8 9 75.91 56.02 42.93 31.91 41.36 20.94 107.40 122.50 40 37.8 28.4 22 5.4 15 16 26 12 16.2 5.8 5.2 9.6 8.2 8 9 31 8.4 9.2 75.13 58.20 42.59 31.74 39.41 21.21 111.53 117.07 4L.8 :) 27 22.2 B 15 17 27 12.2 16.6 6.2 5.2 9.8 8.6 7.4 8.4 31.4 8.2 9 69.23 56.92 42.56 31.28 38.46 18.97 119.34 119.23 Basilar length of Hensel 55vomatic breadth Mastoid breadth 32.4 7.2 21 23 35.4 10 23.4 6 7.8 11.6 11 10 11 37.2 9 10.4 78.00 64.80 46.80 32.00 42.00 20.00 76.92 105.45 29.8 7.8 18.2 21 33.2 15 21.5 6.4 7.8 11.6 11.2 8.6 10.4 42.4 9.4 11 62.55 46.53 32.46 39.39 18.61 82.05 103.57 Interorbital constriction Greatest length of nasals Incisor to molar (alveoli) Incisor to post-palatal notch . . . Height of crown from inferior lip of foramen magnum Fronto -palatal depth at middle of molar series Height of foramen magnum . . . Width of foramen magnum. . . . Greatest length of basi-occipital. Greatest width of basi-occipital. Length of maxillary molar se- ries on crowns .... Length of maxillary molar se- ries on alveoli Greatest length of mandible. . . Length of mandibular molar series on crowns Length of mandibular molar series on alveoli Ratios to basilar length of Hensel Of zygomatic breadth . . . Of mastoid breadth Of fronto-palatal depth . . Of occipital depth Of nasal bones Of maxillary molar se- ries (crowns) Foramen magnum : ratio of height to width Basi-occipital: ratio of length to breadth * VOL. VIM, pp. 117-120 DECEMBER 21, 1893 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON. DESCRIPTION OF A NEW SPECIES OF LAGOMYS FROM ALASKA.* BY E. W. NELSON. Lagomys collaris sp. nov. THE COLLARED PIKA. Lagomys princeps Nelson and True, Report upon Natural History Col- lections in Alaska, pp. 273, 274, 1887 (not Richardson). Type collected about 200 miles south of Fort Yukon, Alaska, near the head of the Tanana river. No. t||f. U. S. National Museum. Col- lected hy E. W. Nelson, 1880. (Original No., 164.) Distribution. Mountains south of Fort Yukon, about the head of the Tanana river, to the Chigmit mountains, near the head of Bristol bay, Alaska. Among the mammals secured by me in Alaska were three specimens of Lagomys which appear to belong to an undescribed species. They were obtained, at my request, through the kind- ness of Mr. L. N. McQuesten, from the mountains south of Fort Yukon, by the Indians of that district. Mr. McQuesten informed me that the Indians report these animals as common everywhere in the highest ranges south of Fort Yukon, where they are usually found above timber line. From native accounts their habits appear to be identical with those of the more southern species. I showed the skins to a fur- trader who had lived for many years at Kolmakovsky Redoubt, * Presented at a meeting of the Biological Society of Washington, December 2, 1893. 18-Bioi,, Soc. WASH., VOL. VIII, 1893. (117) ' 118 Nelson Description of a New Species of Lagonnj*. 011 the Kuskoquim river ; he recognized them at once, and told me that the species is common in all of the high mountains of the Alaskan range south of the Kuskoquim as far as the base of the peninsula of Alaska. In confirmation of this reported range are two skins, now in the United States National Museum, taken in the winter of 1882, in the Chigmit mountains near the head of Bristol bay, by Mr. McKay, which, though in shabby condition, appear to belong to the same species. The three skins secured by Mr. McQuesten were taken by the Indians dur- ing a summer hunt, and are presumably in summer pelage, although possibly taken in spring, before the molt. The Chigmit Mountain specimens are in winter fur, and are much more ashy than those from the type locality. The species thus has a known range from the high mountains south of Fort Yukon, in about latitude 65, southwesterly to the vicinity of Bristol bay. Its northeast limit along this line coin- cides with the same limit of Mazama montana, and here occupies the southern part of the range of Ovis dalli. Its eastern exten- sion remains a matter of uncertainty, but the type locality of Richardson's L. princeps makes it probable that collaris does not reach eastward to the mountains about the head-waters of the Mackenzie. That it does not range north of the Kuskoquim, along the course of that stream, was pretty definitely determined by my work in that region. Description. The dorsal surface, including top and sides of the head, is of a nearly uniform dark or grizzled-gray, with a dull yellowish wash on the crown and back. On the back and sides of the neck the yellowish wash is nearly or quite lacking, leaving a broad collar of dull iron-gray separating the yellowish- shaded areas of the head and back. This yellowish wash is much more apparent on the crown and middle of back and fades out at the sides, so that next the border of the white lower surface the color becomes ashy-gray. Just behind each ear is a small area of dull, light ashy. Below, a triangular white area occupies the entire chin and throat, with its apex in front. At each of the two posterior angles of this area is a yellowish spot forming a slight backward continuation of the light area of the throat at these points. These spots lie on the sides of the throat below and a little behind the ears. The white-throat area is well defined, and is succeeded by a distinct band of dull gray, which forms the lower part of the cervical collar. The remainder of Description of a New Species of Lagomys. 119 the lower parts are pure white, including the feet and legs, except only the smoky-brown fur on the soles of the hind feet. The fore feet and legs to well above their insertion are included within the pure-white area of the lower surface. The exposed portion of the interior of the ear-conch is covered with a strong growth of coarse hairs. These hairs are pale gray at base, shad- ing into yellowish toward the distal third, and the remainder of tip shiny black. The distribution of the hairs is such that the colors form a distinct yellowish band across the middle of the ear, succeeded by a very distinct black bar that extends along the edge and, at first glance, appears to form a black band bordering the ear. A close examination, however, shows that behind this black bar the actual margin of the ear is very nar- rowly edged with pale grayish-white. The black bar is about o mm. in width, and is distinct in all the specimens examined. The under fur of the dorsal surface is slaty -black. Above this slaty-black area the longer hairs have a narrow band of pale gray shading into a broader zone of dull yellowish, followed by a distinct black tip. The two winter specimens from the Chigmit mountains are without any of the yellowish wash of the skins from the type locality, and the long hairs are mainly pale ashy with faint blackish tips, so that the color is a pale ash-gray. The still paler collar separating the uniform color of head and back is present, as is the dull-gray band across the neck below ; otherwise they are colored similarly to the type specimen. All of the specimens have very long pelage as compared with the southern species, and this is especially noticeable in the three specimens from south of Fort Yukon, on which the long dorsal hairs reach an average length of over 22 mm. Measurements. The following average measurements in milli- meters of five dried skins are given merely to serve as an approxi- mation to the true dimensions: Total length, 182; hind foot, 28.3; ear, 19.1 (measured from anterior base). Compared with L. princeps from Idaho and L. schisticeps from the Sierra Nevada of California, the following are the most strik- ing external differences : The hairy, strongly marked ears of the Alaskan animal are conspicuously different from the two south- ern species, both of which latter have the ears thinly covered in front with very short hairs and a much more conspicuous whitish border, On both schisticeps and princeps the yellowish or fulvous 120 Nelson Description of a New Species of Layomys. wash of the upper surface becomes most distinct along the sides bordering the white lower surface, while in collar is it is absent in this region and is most conspicuous on the upper surface of head and back. There also appears to be a higher upward ex- tension of the lower white area along the sides of the latter species. The plain gray of the collar below on the Alaskan animal is replaced on the other species by a band of fulvous or 3^ellowish, which is inclined to spread over the adjoining parts and commonly shades nearly or quite all of the lower surface. There is no sign of this in any of the five specimens of collaris examined by me. The uniform coloration of the upper surface of the head and back, separated by a differently colored collar, is another marked characteristic of collaris. Contrasted with L. princeps and L. schisticeps, the most con- spicuous cranial Character of collaris is the much larger size of its audital bullae. They appear to be larger in every dimension than in the other two species. The post-palatal notch is broad, as in schisticeps, but the pala- tine bridge is broad and heavy and slightly concave on both borders. The interorbital width is greater than in schisticeps. The infra-condylar notch of the mandible is more deeply exca- vated than in either of the other species. In addition, the angular process is much more strongly upturned and ends in a sharp point rising vertically behind the concavity of the notch. The horizontal ramus of the mandible is slenderer than in schisticeps and much as in princeps from Idaho. To the kindness of Mr. F. W. True, Curator of Mammals in the United States National Museum, I am indebted for the opportunity to examine the material on which the present paper is based. Skull measurements of three specimens of Lagomys collaris from the type locality, two hundred miles south of Fort Yukon. U S National Museum number 36297 36298 36296 Basilar length of Hensel ? 35 5 > Greatest zygomatic breadth 21 5 21 5 Interorbital constriction . . . 5 25 5 5 5 5 VOL. VIII, PP. 121-128 DECEMBER 21, 1893 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON, JUNCUS MARGINATUS AND ITS VARIETIES * BY FREDERICK VERNON COVILLE. Jtmcus marginatus has long been known as a species well dis- tinguished from all others, but widely variable within its own limits. Writers on American botany in the early part of the present century gave different names to various forms of the plant, and not until 1866, when Dr. Engelmann published the first part of his Revision of the North American species of the genus Juncus, were they comprehensively treated as forms of Juncus marginatus. The examination of the literature of Juncus marginatus, to- gether with a large series of specimens, has brought the writer to separate the species into three forms, namely, Juncus mar- ginatus (type form), Juncus marginatus aristulatus, and Juncus marginatus setosus, which may be presented as follows : Juncus marginatus Rostk. Juncus marginatus Rostk. Monog. June. 38, t. 2, f. 3 (1801). Type specimen from Pennsylvania. Juncus cylindricus Curtis, Amer. Journ. Sci. xliv. 83 (1843). Type specimen from Lincolnton, North Carolina. Juncus marginatus vulgaris Engelm. Trans. St. Louis Acad. ii. 455 (1866). Type locality not given. * Presented at a meeting of the Biological Society of Washington, De- cember 2, 1893. 19-Bior,. Soc. WASH., Vor,. VIII, 1893. (121) 122 Coville Juncm marginatus and its Varieties. Juncus inarginatus paucicapitatus Engelin. Trans. St. Louis Acad. ii. 455 (1866). Type specimen from Long Branch, Now Jersey. Stems in close tufts, seldom exceeding 50 cm. in height; leaf blades 1 to 2.5 mm. broad ; inflorescence commonly bearing 5 to 15 heads ; heads usually 5- to 10-flowered ; inner perianth parts broadly obtuse ; seed narrowly oblong, 0.4 to 0.5 mm. in length, short-stalked at the base, attenuate-apiculate at the apex, marked with 12 to 16 conspicuous longitudinal costae, the intercostal spaces minutely and closely translineolate. usually with an occa- sional transverse line similar to the cost;r. Specimens examined : Ontario: Near Sandwich, J. M. Bigelow, September 2, 1866. Maine : York, M. L. Fernald, July 16, 1891 ; Cumberland, J. Blake. New Hampshire : East Jaffrey, Walter Deane, July 17, 1889. Massachusetts : Ipswich, William Oakes ; South Frarningham, E. L. Sturtevant, 1890; Cambridge, M. L. Fernald, September 11, 1891 ; Chelsea Beach Island, William Boott, September 10, 1 853. Connecticut: Plainville, J. N. Bishop, 1888. Rhode Island : Providence, George Thurber, June, 1845. New York: Mariners Harbor, Staten Island, Arthur Hollick, September 5, 1886 ; Pine Plains, Lyman Hoysradt. New Jersey : Long Branch, C. W. Short, 1856. Pennsylvania: Lancaster county, T. C. Porter, July 31, 1886; German- town, Thomas Meehan, 1867 ; Grays Ferry, Philadelphia, C. E. Smith ; Wysox, John Carey, 1841 ; Westmoreland county, P. E. Pier- ron, August 11, 1866; Chester county, W. M. Canby, July, 1866; Tinicurn, Bucks county, T. C. Palmer, August, 1892. Delaware : W. M. Canby ; Newcastle, Alexander Commons, 1866. Illinois : Athens, Elihu Hall, 1864 ; Beardstown, C. E. Geyer. Kentucky : Lexington, C. W. Short, 1835. Missouri: Carter county, B. F. Bush, July 21, 1891. District of Columbia : Near Washington, L. F. Ward, July 9, 1876, June 23, 1878, July 21, 1878, and June 14, 1879 ; Washington, M. S. Bebb, 1864. Virginia: Between Princess Anne and Berkeley, A. A. Heller, Xo. 1074, July 13, 1893; along New River, Carroll county, altitude, 2,200 feet, J. K. Small, July 12, 1892 ; Colonial Beach, F. V. Coville, July 6, 1890; Ocean View, Norfolk county, F. V. Coville, June 21, 1890, North Carolina : Lincolnton, Lincoln county, M. A. Curtis; near Salis- bury, Rowan county, A. A. Heller, June 21, 1890; Carolina Beach, New Hanover county, F. V. Coville, June 27, 1890. Florida : A. W. Chapman ; Duval county, A. H. Curtiss, No. 2975 in part. Juncus marginatus and its Varieties. 123 Juncus marginatus aristulatus (Mx.). Juncus aristulatus MX. Fl. i. 192 (1803). Type specimen from either the Carolinas or Georgia. Juncus aristatus Pers. Syn. PL i. 385 (1805). Type specimen the same as that of Juncus aristulatus. Juncus biflorus Ell. Bot. i. 407 (1817). Type locality, ten miles from Savannah, on the road to Augusta, Georgia. Juncus marginatus odoratus Torr. Fl. Nor. U. S. i. 362 (1824). Type specimen from " Bloomingdale, near New York." Juncus heteranthos Nutt. Trans. Amer. Phil. Soc. new ser. v. 153 (1832-37). Type specimen collected along the Arkansas River. Juncus canaliculatus Liebm. Mex. June. 43 (1850). Type specimen collected near San Antonio Huatusco, in the state of Vera Cruz. Mexico, at the altitude of 1,370 meters. Juncus odoratus Steud. Syn. PL Glum. ii. 304 (1855). Type locality the same as that of Juncus marginatus odoratus. Juncus marginatus biflorus Engelm. Trans. St. Louis Acad. ii. 455 (J866). Type locality the same as that of Juncus biflorus. Stems single or in loose tufts, usually exceeding 50 cm. in length ; leaf blades 1 to 5 mm. broad ; inflorescence usually bear- ing 20 to 100 heads, in depauperate plants sometimes fewer ; heads usually 2- to 5-flowered, becoming occasionally 10-flow- ered ; inner perianth parts broadly obtuse ; seed as in the type form, but 0.5 to 0.6 mm. long. Specimens examined : New York : New York, at Bloomingdale, John Torrey. New Jersey : Near Atsion, C. F. Parker, July 31, 1866 ; near Princeton, John Torrey, July, 1830 ; Burlington county, C. F. Parker, August 19, 1867; Dennisville, Cape May county, C. F. Parker, July 15, 1866. Pennsylvania: West Chester, William Darlington, 1827. Delaware : Ellendale, W. M. Canby, September 16, 1891 ; near Wilming- ton, Alexander Commons, July 7 and 24, 1866. Maryland: Garrett county, J. D. Smith, July, 1879; Salisbury, Alex- ander Commons, July 27, 1865 ; Salisbury, W. M. Canby, June, 1864. Virginia: Virginia Beach, Princess Anne county, Arthur Hollick and N. L. Britton, September 26 and 27, 1890 ; near Virginia Beach, A. A. Heller, No. 1053, July 12, 1893; Colonial Beach, F. V. Coville, July 6, 1890; Ocean View, Norfolk county, F. V. Coville, June 21, 1890. District of Columbia: Near Washington, Lester F. Ward, July 21, 1878. North Carolina: Wilmington, M. A. Curtis; near Wilmington, F. V, Coville, June 26, 1890. 124 Coville Juncus marginatus and its Varieties. South Carolina : Aiken, H. W. Ravenel, June 12 and 25, 1866, and August, 1869. Florida : Duval county, A. H. Curtiss, No. 2975, in part ; Fort Meade, Polk county, J. D. Smith, March, 1880; Gadden county, A. W. Chapman. Alabama: Alexander Winchell, No. 162; Montgomery, Gerald McCarthy, 1888 ; Fly Creek, Baldwin county, Charles Mohr, June 13, 1890. Mississippi: Starkville, S. M. Tracy, Nos. 1420, 1421, June 20, 1890 ; Ocean Springs, S. M. Tracy, No. 1648, July 10, 1891. Louisiana: Josiah Hale. Texas: Hardin county, G. C. Nealley, 1888; Houston, F. Lindheimer, 1842. Michigan : Near Fort Wayne, J. M. Bigelow, July 28 to September 9, 1866. Kentucky : Lexington, C. W. Short, 1835. Kansas : Cherokee county, W. S. Newlon, 1893. Missouri : St. Louis, C. A. Geyer, May, 1842. Arkansas: F. L. Harvey, 1882 to 1884; Little Rock, H. E. Hasse, May 25, 1886. Indian Territory : Choctaw agency, J. M. Bigelow. Guatemala: Coban, Alta Vera Paz, altitude, 4,300 to 4,400 feet, H. von Turckheim, No. 431, April, 1886, and April, 1879. Brazil: " Brasilia meridional!," Sellow. Juncus marginatus setosus var. nov. Stems apparently in loose tufts, 30 to 75 cm. high ; leaf blades 1 to 5 mm. broad ; inflorescence and heads as described in Juncus marginatus aristalatus ; inner perianth parts narrowly ovate to lanceolate, 0.3 to 0.4 mm. in length, reticulated in 12 to 16 longi- tudinal rows, the areolse nearly isodiametrical, transversely pluri- lineolate. Type specimen in the United States National Herbarium, col- lected June 4, 1882, in the Santa Catalina mountains, Arizona, by C. G. Pringle. Specimens examined : Kansas : Stafford and Kingman counties, M. A. Carlton, 1891. Nebraska: Minderi, Kearney county, June, 1891, intergrading with the type form. Arizona: Santa Catalina mountains, C. G. Pringle, June 4, 1882, and April 14 and May 16, 1881 ; Lowell, W. F. Parish, May 24, 1884 ; Apache Pass, J. G. Lemmon, No. 313, 1881. New Mexico: Charles Wright, No. 1923 in part, 1851. Indian Territory: Colbert, C. S. Sheldon, No. 37, June 19, 1891 ; Russell Creek, W, S, Newlon, 1893. Juncus marginatus and its Varieties. 125 Texas: Brummond, F. Lindheiiner, No. 193, 1843; Corpus Christi, H. W. Ravenel, April 30, 1869; Houston, Elihu Hall, PI. Tex. No. 657, April 20, 1872. Arkansas: Little Rock, F. L. Harvey, June, 1880. Louisiana: Port Eads, on ballast ground, A. B. Langlois, May 6, 1885. Mexico : Near Morales, San Luis Potosi, J. G. Schaffner, No. 500, 1876 ; Rio Blanco, Jalisco, Edward Palmer, No. 13, June 8, 1886. The considerations which have led to the disposition of the species formulated above may be given in sufficient detail to aid the future student who goes over the same ground. The type specimen of Juncus marginatus, which is probably at Berlin, it has not been possible to consult, but Rostkovius's ex- cellent figure, as well as his description of the plant, u culmus erectus pedalis vel sesquipedalis," " corymbus terminalis simplex," and " capitula octo- vel decemflora" besides the type locality, Penn- sylvania, leave no doubt as to the identity of the type form. The type specimen of June as cylindricus, a fragment of which I have been able to examine in the Engelmann herbarium, bears heads with abnormally elongated axes and many flowers. It belongs clearly to some form of Juncus marginatus, probably to the type form. Dr. Engelmann, following a practice in com- mon European use, gave to what he considered the type form a varietal name, vulgaris, which, by reference to No. 33 of his Herbarium Normale, is seen clearly to be identical with Rost- kovius's plant. An examination of the type specimen of Juncus marginatus paucicapitatus in the Engelmann herbarium shows that it belongs to the type form of Juncus marginatus, and indeed closely resembles Rostkovius's original figure of the species. The plant for which the name Juncus marginatus aristulatus is here adopted has been known currently as Juncus marginatus biflorus. Michaux's Juncus aristulatus has been referred by vari- ous authors to Juncus marginatus without particular comparison with any of its forms, but an examination of his description leaves no doubt regarding the plant he was describing, for in the expression " glomerulis trifioris " he names the most con- spicuous external feature of this variety. Persoon's specific name aristatus is an error for aristulatus, for Persoon cited Michaux as the author of the name, with reference to the page of publica- tion, and quoted his description with but slight changes. Re- garding Juncus biflarus, it should be said that Elliott included in his book descriptions of two plants of the marginatus group, 126 Coville Juncus marginatus and its Varieties. one, which he referred to Juncus aristulatus MX. (citing it by error, apparently from Pursh, Juncus aristatus. and modifying Michaux's description to some extent), and another, which he described as a new species, Juncus biflorus, with the diagnosis ' Juncus culmo tripedali, tereti ; foliis linear ib us y planis ; panicula decomposita, elongata ; glomerulis bifloris" There can be no doubt that this plant is referable to true Juncus aristulatus, and that Elliott incorrectly transferred Michaux's name to some other form, perhaps the type form, of Juncus marginatus. Torrey's Juncus marginatus odoratus, the original specimen of which I have examined in the Columbia College herbarium, is Juncus marginatus aristulatus, with 3- to 5-flowered heads, and has now lost the pleasant odor, probably accidental, attributed to it by Dr. Torrey. Nuttall's Juncus heteranthos can be nothing else than Juncus marginatus biflorus, for although his type specimen appears to have been lost he states explicitly in his description that the flowers are mostly in threes, and that the inner perianth parts are obtuse. In describing the plant, therefore, he appears to have been distinguishing it from the common typical eastern form of Juncus marginatus and not to have had in mind Elliott's Juncus biflorus. The form ascribed to the inner perianth parts precktdes its reference to Juncus marginatus setosus. The type specimen of Liebmann's Juncus canal iculat as has been examined by Dr. Franz Buchenau, who refers it unhesitatingly to Juncus marginatus* Since he does not say that its inner perianth parts are acute, I judge that it is not Juncus marginatus setosus, which, indeed, is very unlikely to occur in the humid climate of Vera Cruz. Neither is Liebmann's plant referable to the type form of Juncus marginatus, for that plant does not range so far south. The Juncus odoratus of Steudel is based on Juncus marginatus odoralus Torr., to which reference has already been made, and Engelm aim's Juncus marginatus biflorus is based upon Juncus biflorus Ell. The variety described above as new could not, from its range, have been referable to any of the plants whose names have been cited above as synonyms of Juncus marginatus or its variety, with the exception of Juncus heteranthos and Juncus canal iculatus, and the reasons for the reference of these two plants to Juncus mar- ginatus aristulatus have been given in the last paragraph. The localities of specimens examined in the preparation of , Naturw. Yer, Bremen, iii. 343-344 (1873), Juncus marginatus and its Varieties. 127 this paper have been taken from the labels in the herbaria of Harvard University, Columbia College, the Missouri Botanical Garden, and the Academy of Natural Sciences of Philadelphia, the private herbaria of Mr. William M. Canby and Captain John Donnell Smith, and the National Herbarium. The disposition which is here made of the type form and varieties of Jiuicus marginatus may serve as a general illustration of a method of dealing with species and varieties which is in accord with our present knowledge of the evolution of species and of the geographic relationship of these plants with each other. The writer, having first secured a considerable amount of material, separated the specimens into the three groups which their examination naturally suggested. Next the names which have been published for any one of these forms were taken jap and the original descriptions and the type specimens examined. In some cases the types were not accessible, and under those circumstances it was necessary to identify the plant either by description alone or by the aid of collateral evidence. In this manner it is believed that the earliest name applied to each of these forms has been ascertained, one of them, as it proved, hav- ing never before received a distinctive name. The next step was to bring together a still larger number of specimens, identifying each one according to the facts already known, ascertaining the locality in which it was collected, and marking its position upon a map. It was found .that the type form of Juncus marginatus merges by a full series of intergrades into Juncus marginatus aristulatus, and that in the states of Nebraska and Kansas it seems to intergrade also with Juncus marginatus setosus. The latter shows a close relationship with the variety aristulatus, but, so far as indi- cated by the specimens examined, does not fully intergrade with it. The differences between these two, however, are so slight that there is reason to expect the occurrence of intergrades. The type form of Juncus marginatus ranges from Maine southward through the Atlantic States to Florida, and occurs again at a point in the province of Ontario opposite Detroit. The variety aristulatus ranges from New York city southward along the coastal plains to Florida, westward through all the States border- ing the Gulf of Mexico, and northward, in the Mississippi valley, in apparently isolated localities, to southern Michigan. South of the United States it follows the eastern coast at least as far as Coban, Guatemala, and one specimen is reported to have 128 Coville Juncus marginatus and its Varieties. been found in southern Brazil. The variety setosus occurs in the southern Great Plains region, ranging over western Kansas, Indian Territory, interior Texas, Arizona, and New Mexico, and southward in Mexico to the states of San Luis Potosi and Jalisco. An isolated locality is also known on ballast ground in Louisiana and another in Arkansas. The ranges of the type form and the variety aristulatus appear to overlap in the Atlantic States over a considerable area, but in reality they are pretty well distinguished, for the latter characteristically inhabits sandy plains, particularly those along the coast, while the type form grows more abundantly in upland moist areas. The most widely diffused and probably the mother form of Juncus marginatus is its variety arixtnly the expedition in Kern Valley, Walker Pass, Owens Valley, ami in the Coso, Argus, and Panamint Mountains, California. Spermophilus chrysodeirus brevicaudus subsp. nov. SAN BERNARDINO SPERMOPHILE. Type from SAN BERNARDINO PEAK, CALIFORNIA. No. 56661 9 ad. United States National Museum, Department of Agriculture collection. Collected October 9, 1893, by J. E. McLellan (original number, 274). Measurements of Type Specimen (taken in flesh). Total length, ; tail vertebrae, ; hind foot, . General Characters. Similar to S. chrysodeirus, but with much shorter tail, somewhat shorter hind foot, and duller mantle over head and shoulders. The tail averages about 75 mm., while that of chrysodeirus averages 90 mm. or more. Color (of type specimen). Back and rump grizzled gray tinged with brownish ; sides paler ; a dull fulvous mantle over head and neck, hardly reaching shoulders ; color of head shading to- ward brick-red ; sides of neck behind ears buffy-ochraceous ; a broad whitish stripe, bordered on each side, by a broad black stripe, extends from the shoulder to the rump on each side, and the white reaches beyond the black in both directions ; hind foot dull whitish : tail above, proximal half grizzled'; distal half black, edged with fulvous ; tail below, chestnut, bordered with black and edged with fulvous. Number of specimens examined, 7 ; all from San Bernardino Mountains, California. Tamias panamintinus sp. nov. PANAMINT CHIPMUNK. Type from PANAMINT MOUNTAINS, CALIFORNIA (between Death Valley and Panamint Valley). Exact locality, Johnson Canon. No. |f8!s C? ad. United States National Museum, Department of Agriculture collec- tion. Collected by E. W. Nelson April 3, 1891 (original number, 723). Measurements of Type Specimen (taken in flesh). Total length, 208 ; tail vertebra, 96 ; hind foot, 31. Ear from notch, 16 (measured in dry skin). Average measurements of 50 adult specimens from type locality : total length, 208; tail vertebrae, 90.4; hind foot, 31. General Characters. About the size of T. quadrivittatiis. but re- sembles T. speciosus much more closely in coloration, being a paler form with obsolescent facial stripes and less distinct dorsal stripes, which are shorter posteriorly, leaving the rump clear gray. Ears smaller than in speciosus; ear stripes nearly obso- New Ground Squirrels. 135 lete. Tail rich -orange-rufous both above and below and only narrowly margined with black. Hind foot shorter than in any other member of the speciosus group. Color. Winter pelage : Top of head and rump ash gray; back of neck and inner pair of dorsal light stripes pale vinaceous drab, fading into gray posteriorly ; dorsal stripes short, not continuing over rump ; outer pair of light stripes dull whitish ; all five dark dorsal stripes dull fulvous ; sides gray, washed with buffy ochra- ceous. Face stripes indistinct; those below the eye obsolete. Tail: upper surface rich orange rufous (from the broad sub-basal zone of this color), which is only heightened by the narrow sub- apical zone of black and the yellowish tips of the hairs ; under surface deep orange rufous, with a narrow submarginal band of black, bordered with yellowish. Summer Pelage : Similar to winter pelage, except that the sides and dark dorsal stripes are suffused with ferruginous, and the median part of the central stripe is blackish posteriorly. Cranial Characters. Skulls of T. panamintinus resemble those of T. quadrivittatus from the type locality in Colorado so closely that the two are hardly distinguishable, though the skull of the Panamint animal is slightly smaller and more depressed in the fronto-nasal region, and has larger audital bullse. Contrasted with T. speciosus, the brain case is flatter and the audital bulla? conspicuously larger. General Remarks. In winter pelage panamintinus differs from speciosus, the only form with which it requires comparison, in having all of the stripes less distinct, particularly those of the face and ears ; the dorsal stripes shorter, not running back over the rump, which is clear gray ; the outer pair of white stripes narrower, shorter, and less pure white; the dark dorsal stripes pale fulvous instead of dark umber ; the shoulders and back of the neck suffused with buffy-ochraceous instead of being gray, and the rufous of the distal half of the tail not obscured by black. In summer pelage it resembles speciosus much more closely, but may be distinguished by less vivid tints, paler facial stripes, narrower outer dorsal white stripe, obsolescent ear stripes, pale gray rump, and by the small amount of black on the tail. In all pelages the black on the tail is very much restricted, permit- ting the rufous to show through on the upper surface for its entire length, thus imparting to it a peculiar ruddy glow not seen in any other species. 136 Merriam New Ground Squirrels. Geographic Distribution. Tamias panamintinus is an inhabitant of the desert ranges of the west side of the Great Basin in Cali- fornia and Nevada, where 110 specimens were obtained by the Death Valley Expedition. Tamias callipeplus sp. nov. MOUNT PINOS CHIPMUNK. Type from summit of MOUNT PINOS, VENTURA COUNTY, CALIFORNIA. Xo. IMff cT yg--acl. United States National Museum, Department of Agri- culture collection. Collected by E. W. Nelson October 20, 1891 (original number, 1344). Measurements of Type Specimen (taken in flesh). Total length, 212 ; tail vertebrae, 92; hind foot, 33.5. Average measurements of four specimens from type locality: total length, 210; tail vertebrae, 91.7 ; hind foot, 34. General Characters. Agrees with speciosus, its nearest relative, in size, proportions, and pattern of markings, including the great breadth of the outer white dorsal stripe. It differs from speciosus in having the thighs and rump yellowish instead of gray, the back of the neck and inner pair of light dorsal stripes vinaceous- drab instead of gray ; the post-auricular patches larger, purer white, and more sharply defined, and the black on the tail much less extensive. Ears large. Color. Winter pelage : No gray anywhere ; top of head, back of neck, and inner pair of light dorsal stripes vinaceous drab, tinged with ochraceous on the shoulders, becoming fulvous on the flanks, and yellowish on the thighs and rump ; outer white stripes very broad (as in speciosus) and slightly obscured poste- riorly by dark-tipped hairs ; median dorsal stripe dark umber- brown, bordered and obscured by rusty ; inner pair of dark stripes ferruginous; outer pair fulvous, not defined below, pass- ing into fulvous of flanks ; post-auricular spots large, sharply defined, and pure white ; ear stripes sharp, the posterior pure white, the anterior black, edged in front basally with rusty ; facial stripes intensely colored and sharply defined, the middle or orbital stripe black, becoming rusty at the base of the ear ; feet faintly washed with fulvous ; tail orange rufous, broadly tipped and narrowly bordered with black, and edged with yel- lowish ; the rufous obscured on upper surface by black sub apical and yellowish apical zones on the hairs ; belly and throat pure white, the dark basal color showing through in places. Cranial and Dental Characters. No cranial characters of im- portance have been discovered, though the brain case is slightly \<-w Ground Squirrels. 137 more arched in the posterior frontal region than in speciosus. The molariform teeth are somewhat heavier, and the last upper molar has the heel more developed. General Remarks. Tamias callipeplus differs from both speciosus and panamintinus in having the hinder parts of the body yel- lowish instead of gray, and in the purer white and larger size of the post-auricular spots. It agrees with panamintinus and differs from speciosus in the vinaceous tinge of the back of the neck and inner pair of pale dorsal stripes, and in the color of the tail. It differs from panamintinus and agrees with speciosus in the bright- ness and sharpness of the facial stripes and ear stripes, the great breadth of the outer pair of dorsal white stripes, in the posterior extension of the dorsal stripes over the rump, and in the large size of the hind foot. Tamias callipeplus is treated as a full species instead of a sub- species on account of its isolated geographic position, intergrades being impossible because the mountains on both sides of Mount Finos do not attain sufficient altitude to provide the cool tem- perature required by the species. Tamias alpinus sp. nov. ALPINE CHIPMUNK. Type from BIG COTTONWOOD MEADOWS, HIGH SIERRA, CALIFORNIA, just south of Mount Whitney (altitude, 3,050 meters or 10,000 feet). No. f f f $ 9 yg.-ad. United States National Museum, Department of Agriculture collection. Collected by Basil Hicks Dutcher August }2, 1891 (original number, 191). Measurements of Type Specimen (taken in flesh). Total length, 189; tail vertebrae, 82 ; hind foot, 29. Ear from notch, 13 (in dry skin). Average measurements of 15 specimens from type locality : total length, 185 ; tail vertebrae, 79; hind foot, 29.3. General Characters. Size, small ; resembles T. minimus pictus in size, proportions, and general appearance, but is much paler in breeding pelage and much more ferruginous in midsummer pelage; may be distinguished from pictus in all pelages by the tail, which is much broader and more bushy, hoary above, and broadly tipped with black both above and below. Color. Fall pelage : General color of upper parts, hoary gray ? suffused on the flanks with buffy-ochraceous ; median dorsal stripe dusky, obscured by pale rusty ; lateral dorsal dark stripe pale ferruginous ; inner pair of white stripes hoary gray ; outer pair white and very broad (as in speciosus); post-auricular 21 BIOI.. Soc. WASH., Vou VIII, 1893. 138 Merriam Ar?/> (1 round ^l patches whitish, not sharply defined; facial stripes pale; ear stripes indistinct ; legs and feet gray. Tail : upper surface hoary (rarely yellowish), becoming black toward the tip (the individual hairs buffy gray aub-basally , then black, and broadly tipped with pale buffy gray or yellowish) ; under surface, pale buffy fulvous, bordered and broadly tipped with black, broadly edged laterally with pale buffy. Summer Pelage : Dorsal dark stripes bright ferruginous ; facial stripes strengthened by dull rusty ; flanks bright fulvous, the fulvous -reach ing forward over shoulders to sides of neck. Cranial Characters. Skull similar in size and general appear- ance to that of T. minimus pictus, but with longer nasals and nasal branches of premaxillaries. The length of the nasals equals or exceeds the combined length of the basioccipital and basisphenoid. In minimus the nasals fall considerably short of this measurement;* General Remarks. The Alpine chipmunk is one of the two smallest chipmunks known, the other being the Sage Plains species (T. minimus), which it resembles in general appearance, except in the full summer pelage. In all pelages it may be dis- tinguished from minimus by the tail, which is hoary above (rarely yellowish) ; is broader and more bushy, and has the black terminal part much longer. The outer pair of white dorsal stripes also are much broader, as in speciosus. In spring and early summer, before the post-breeding molt, the animal is very much paler than the palest specimens of minimus pictus. In midsumer pelage, on the other hand, the sides and dark stripes are deeper ferruginous than ever seen in the brightest summer specimens of minimus pictus or even minimus consobrinus, and in high-colored individuals even the inner pair of light stripes are sometimes obscured by rusty. Geographic Distribution. This beautiful little chipmunk is re- stricted, so far as known, to the alpine summits of the High Sierra, where it lives among rocks at timber-line, ranging a little above and a little below the upper limit of tree growth. Thus the haunts of the alpine chipmunk are the same as those of the pika (Lagomys), the alpine marmot (Arctomys flaviventer), and the mountain sheep (Ovis canadensis). No mammal ranges higher. Sixty specimens were obtained by the Death Valley Expedition. *This has been verified in 100 skulls, 50 of alpinus and 50 of niininmx and subspecies. VOL. VIII, PP. 139-142 DECEMBER 28, 1893 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON. DESCRIPTION OF A NEW SPECIES OF ARVICOLA, OF THE MYNOMES GROUP, FROM ALASKA. BY E. W. NELSON. Arvicola operarius sp. nov. THE TUNDRA MOUSE. Anicoln r!jntrinx homdis Nelson and True, Report upon Natural History Collections in Alaska, 1887, pp. 275, 276, Series A (not Richardson). Type No. gfall- United States National Museum. From St. Michaels, Norton Sound, Alaska. November, 1879. Collected by E. W. Nelson. Measurements. Average measurements, in millimeters, of five dried skins, fall and winter specimens, from St. Michaels are as follows: Total length, 104.5; length of tail vertebra?, 25.2; hind foot, 17.9 ; ear from front base, 9.8. Color. The dorsal surface, including the top and sides of the head, is pale, dull fulvous or fawn color, thinly washed with darker from the overlying black tips of the long hairs. On the sides the fulvous of the back shades gradually into the paler lower surface and sometimes forms a faint wash over most of the under parts. The top of the tail is dark brown, in contrast with the color of the back, and its sides and lower surface are white. The under surface is plain, dull, grayish white, including upper lips, chin, and throat T and extending laterally to the insertion of the legs. In some cases the white area is separated from the white under surface of the tail by a narrow band of fulvous 22-BiOL. Soc. WASH., VOL. VIII, 1893. (139) 140 Nelson A New Species of Arvicola. which incloses the base of this member below. The feet and legs are dull whitish. The fur is long and very soft, except on the tail. The tail hairs are rather coarse and stiff and the termi- nal pencil is peculiarly bristly with a length of 6 mm. The ears are rounded and clothed on the inner side of conch by yellow hairs. They are wholly concealed by the fur. The under fur, with basal two-thirds of the long hairs, is of a uniform dark slaty color ; succeeding the dark base the majority of the long hairs of the back have a plain fulvous tip. The tips of the longest hairs, however, are black, succeeding a yellowish zone, thus pro- ducing a slight wash of darker over the back. A slight variation in color is apparent among the specimens, due to the varying intensity of the fulvous. The soles of feet are naked. The thumb nail is short, stout, and spatulate and does not extend beyond end of the thumb. Dentition. The accompanying figure shows the pattern of the crowns of the molar teeth. Skull measurements of five specimens of Arvicola operariu* from St. Michaels, Alaska : Molar teeth of Arvi- cola operarius x 5. a, upper; 6, lower. U. S. National Museum numbers. . 22225 22212 22224 22214 22223 Basilar length of Hensel 21 21 21 21 20 75 Greatest zygoma-tic width 12.5 12.5 12.5 12.75 12.5 Interorbital constriction 3.5 3 . 5 3.5 3 . 25 3.5 The present species may be readily distinguished from any known American Arvicola by its smaller size and pale fulvous color. It is very abundant along the coast tundras of Bering Sea from Cape Vancouver north at least to Bering Straits, in- cluding Nelson, St. Michaels, and Stewart Islands. It makes numerous runways through the moss and under the grass in all the tundra districts where it lives, and as winter approaches gathers stores of small bulbous roots, sometimes placing a peck or more in a single cavity just below the surface on a mossy knoll or slope. A New Species of Arvicola. 141 For a short period before the first snowfall in autumn the Eskimo women and children search for these stores with pointed sticks, which they thrust into the sides of mossy banks in suit- able places, the spot being found by the ease with which the stick penetrates the few inches of mossy cover. In this way con- siderable quantities of this food are gathered, and during the following winter it is boiled and eaten as a delicacy. The boiled roots have a flavor like a boiled unripe sweet potato and are very palatable during the long winter fare of meat and fish. During seasons when the snow remains on the ground from fall until spring, comparatively few of these mice come about the houses until the snow begins to melt in spring, when they always become numerous there. A winter thaw occurs at intervals of several years, melting away nearly all of the snow. At such times the water perco- lates into all of their runways and storehouses, and the quickly succeeding cold freezes them solid for the remainder of the season. In this way the majority of these mice are at once be- reft of shelter and food, and are found wandering about on the surface of the tundras, where many are eaten by foxes and other animals, while others freeze to death, and scores swarm about the native villages and the fur-trader stations. Ordinarily in spring, as the snow melts away, many winter burrows are revealed just at the lower surface of the snow. Their burrowing can never extend very deep in many places where the permanently frozen soil lies at a depth of from one to two feet. On a dry peat knoll fronting the sea near St. Michaels I once followed their holes to a depth of about two feet. The Eskimo boys trap them in toy traps modeled after those used by the men for larger game, and the children use their skins for blankets and clothing for dolls. These mice are omnivorous, and when two or more are con- fined together in a box the stronger usually kill and partly devour the weaker. Through the kindness of Dr. C. Hart Merriam I am able to present herewith a plate showing the character of the dentition of this species. I wish also to acknowledge herein my indebted- ness to the courtesy of Mr. F. W. True, Curator of Mammals in the United States National Museum, who kindly placed the speci- mens of the Alas-kan Arvicolte at my disposal. VOL. VIII, PP. 143-146 DECEMBER 29, 1893 PROCEEDINGS OF THB BIOLOGICAL SOCIETY OF WASHINGTON PRELIMINARY DESCRIPTIONS OF FOUR NEW MAMMALS FROM SOUTHERN MEXICO, COLLECTED BY E. W. NELSON. BY C. HART MERRIAM, M. D. Lepus orizabae sp. nov. MT. ORIZABA COTTONTAIL. Type from MT. ORIZABA, STATE OF PUEBLA, MEXICO (altitude about 9, 500 feet). No. 53318 9 ad. United States National Museum, Department of Agriculture collection. Collected April 24, 1893, by E. W. Nelson (origi- nal number, 4730). Measurements of Type Specimen (taken in flesh). Total length, 395; tail vertebrae, 51 ; hind foot, 90. General Characters. Similar to L. arizonaR in general appear- ance and length of ear, but very much darker and differing con- siderably in color ; pectoral band broad and full, forming a dis- tinct ruff. Color. Head, back-saddle, and thigh streaks grizzled buffy clay-color and black, the admixture of black plentiful and uni- form ; flanks and rump mixed gray and black, the gray particu- larly clear on rump ; upper side of tail grizzled drab-gray, with tips of hairs buffy ; nape patch and fore and hind feet, dull 23-BiOL. Soc. WASH., VOL. VIII, 1893, (H3) 144 Merriam New Mammals from Southern Mexico. fulvous, palest on inner sides of hind feet ; pectoral collar drab, with tips of long hairs buffy ; chin, throat, middle of belly and under side of tail, pure white with under fur plumbeous except under tail, where the fur is white throughout; ears dull grayish brown, bordered apically with a brown band, which is not very well defined ; anterior edge fringed with whitish. Cranial Characters. Skull similar to that of L. arizonze, but with audital bulla3 much smaller. Sciurus nelsoni sp. nov. NELSON'S SQUIRREL. Type from HUITZILAC, MORELOS, MEXICO. No. 51157 9 ad. United States National Museum, Department of Agriculture collection. Collected January I, 1893, by E. W. Nelson (original number, 4144). Measurements pf Type Specimen (taken in flesh). Total length, 500; tail vertebrae, 246 ; hind foot, 68. Average measurements of 4 specimens from type locality : Total length, 527 ; tail vertebrae, 262 ; hind foot, 70. General Characters. Premolars f- ; size large, about equaling S. cervicalis; color of upper parts uniform. No trace of nuchal patch or rump patch. Color. Type Specimen : Upper parts from nose to tail grizzled yellowish brown; under parts grizzled fulvous, purest on throat and breast, much mixed with black posteriorly ; end of nose, ears, and feet black, the feet (both fore and hind) more or less grizzled (probably a seasonal character, as the feet are wholly black in most of the specimens); tail black; upper side with tips of hairs whitish ; under side bordered with whitish. Below the broad subterrninal black zone there is a concealed zone of dull fulvous. Other Specimens. Some specimens are much darker than others, and the difference seems to be seasonal. In some the head, feet, and legs are intense black. The feet are almost wholly black in nearly all the 14 specimens, and the under parts are usually blacker than in the type. The tail is usually black, washed above and on the sides with whitish, but in some speci- mens the under side is grizzled from admixture of black and dull buffy -gray. Two or three of the specimens show a whitish spot (usually rather indistinct) at the posterior base of the ear. Specimens examined, 14 : 4 from Huitzilac, Morelos ; 9 from Ajusco, Mexico, and 1 from Salazar, Mexico, New Mammals from Southern Mexico. 145 Thomomys orizabae sp. nov. MT. ORIZABA THOMOMYS. Type from MT. ORIZABA, STATE OF PUEBLA, MEXICO (altitude, about 9,500 feet). No. 53616 ad. United States National Museum, Depart- ment of Agriculture collection. Collected April 25, 1893, by E. W. Nelson (original number, 4744). Measurements (taken in flesh). Type specimen: Total length, 217 ; tail vertebrae, 68; hind foot, 30. Average measurements of 13 specimens from type locality : total length, 213 ; tail vertebrae, 66 ; hind foot, 29. General Characters. Size medium ; sooty-plumbeous phase dominant ; fulvous phase resembling T. fulvus, but duller. Tail longer than in T.falvus; well haired; fore and hind feet well haired. Color. Plumbeous phase (the type and 15 out of a total of 17 specimens from type locality are in this phase) : Everywhere uniform slate-black (faintly paler below) except distal part of fore and hind feet, distal third of tail, and inside of cheek pouches, which- are white (sometimes also a few white hairs about mouth and under chin). The color of the body always passes down over the wrists and ankles and usually reaches half way to the toes--sometimes further. Fulvous phase: Upper parts dark umber-brown, becoming dusky on nose and dull fulvous on sides ; under parts buffy- fulvous, the plumbeous basal fur showing through in places ; under side of face blackish ; feet and distal third of tail white. (Only 2 specimens, one of which is very young, out of a total of 17, are in this pelage.) Cranial Characters. The skull of T. orizabse differs from that of T. peregrinus here described (the only species thus far recorded from southern Mexico) in the following particulars : Muzzle longer and much broader ; frontals anteriorly much broader ; ascending branches of premaxillae much broader and blunter posteriorly. The breadth of muzzle across ascending^branches of premaxill&e, and breadth of frontals anteriorly, is considerably greater than the interorbital breadth ; in peregrinus the contrary is true. 146 Merriam New Mammals from Southern Mexico. Thomomys peregrinus sp. nov. WANDERING THOMOMYS. Type from SALAZAR, STATE OF MEXICO. No. 50130 $ yg.-ad. United States National Museum, Department of Agriculture Collection. Collected October 23, 18!)2, by E. W. Nelson (original number, 3668). Measurements (taken in flesh.) Type specimen : Total length, 207 ; tail vertebrae, 72 ; hind foot, 28.5. Average measurements of 10 specimens from type locality : Total length, 201 ; tail vertebrae, 57 ; hind foot, 28. General Characters. Similar to T. fulvus, but much darker and duller, lacking the bright golden-fulvous tints of that species, and differing also in cranial characters. The incisors curve far forward. Color. Upper parts dark^ umber-brown, becoming sooty-black on head and along middle of back, and dull fulvous on sides ; end of nose, broad ring round mouth, and ear-patches blackish ; under parts buffy, the plumbeous basal fur showing through ; toes of fore feet, most of hind feet, and distal i or of tail white. On the hind feet the dark color of upper parts reaches down over ankles ; on fore feet it usually comes down to or near bases of toes. Cranial Characters. Compared with T. orizabae the muzzle is very narrow, the breadth across ascending branches of premax- illse being much less than the interorbital breadth. The frontals anteriorly are also considerably narrower than interorbitally, and the premaxillaB are slender and pointed posteriorly. In young and middle-aged skulls the interparietal is straight pos- teriorly, broadly rounded anteriorly, and nearly twice as broad as long ; in old skulls it is roughly subquadrate. VOL. VIII, pp. 147-152 DECEMBER 29, 1893 PROCEEDINGS OF THE BIOLOGICAL SOCIETY OF WASHINGTON. THE YELLOW BEAR OF LOUISIANA, URSUS LUTEOLUS GRIFFITH. BY C. HART MERRIAM, M. D. Ursm Meolus Griffith, Order Carnivora, 1821, 236-237, with colored plate. Ursus americanus Baird, Mammals of North America, 1857, 217 and 222 (in part not of Pallas V Urxns cinnamomeus Brown, Forest and Stream, December 16, 1893, 519 (in part not cinnamomum of Audubon and Bachman). In the year 1800 Shaw mentioned and in part described a yellow bear from the southern United States, but did not give it a specific name, referring it to the black bear as a variety, with a query. His account in full is , as follows : " Yellow bear. Among Mr. Catton's figures of quadrupeds a representation is given of a yellow bear from the living animal then kept in the Tower. The following is the description accompanying the plate referred to: 'The yellow bear from Carolina (as the American bears in general are) is rather smaller than the European bears ; it has also a more pleasant and agreeable countenance, is perfectly tame and sociable ; the color a lively, bright orange, of a red- dish cast ; the hair thick, long, and silky. Its. other properties are the same as of the species in general.' " (General Zoology, vol. I, part II, Mammalia, 1800, p. 454.) In 1821 Edward Griffith, in his important and rather rare work on the Carnivora, named this bear Ursiis luteolus, and gave 24 RIOT,. Soc. WASH., VOT,. VIII, 1893. (147) 148 Merriam The Yellow Bear of Louisiana. a colored plate of it. His description is as follows : " Yellow bear. Ursus luteolus. The American yellow bear has been spoken of as a variety of the black bear of that continent. Independ- ently of the individual mentioned in Shaw's Zoology, Major Smith has a drawing of one. taken in Louisiana,* and there is a fine specimen now in the Tower, which is aptly called, from its color, the cinnamon bear. This last is smaller; the forehead more convex ; the nose more conical than in the black species ; the ears also stand farther back ; the physiognomy may be said to be more fox-like, and the hair is not so long or thick. It is gentle in disposition, which, indeed, is expressed in the counte- nance of the animal very decidedly. We cannot, therefore, but conclude that the hereditary distinctive differences of color, organization, and moral character are quite sufficient to consti- tute this a separate species. "The yellow bear was formerly common in Virginia, and is still frequently met with in northwestern Louisiana, where it is called the white bear, and seems generally, though without doubt erroneously, to be considered an accidental variety, the offspring of the black bear. It subsists on honey, acorns, &c., as well as flesh." (Descriptions of Vertebrated Animals, Order Carnivora, London, 1821, 236-237, and col. pi.) Six years later, in the mammal part of his well-known edition of Cuvier's Animal Kingdom, Griffith reluctantly treats the spe- cies as a variety of the American black bear, saying : " The Baron [Cuvier] also thinks that the yellow bear of Carolina is a variety of the same species. This is scientifically termed the Ursus luteolus. We shall not venture to assert, in contradiction to the authority of the Baron, that this bear forms a distinct species, but assuredly it is a very strongly marked variety. * * * They were formerly common in Virginia, and they are still abundant in northwestern Louisiana, where they are called white bears, and are said to feed chiefly on honey, on acorns of a large size, wild berries, &c." (Griffith's Cuvier, Mam- malia, II, 1827, 228-229.) Whether or not two distinct bears were confounded in the original description is of little conse- * Lest any one should suppose that the old Territory of Louisiana, stretching westward to the Rocky mountains, was meant, it may be stated that the present boundaries of Louisiana were fixed in 1812, nine years before the publication of Griffith's original description and fifteen years before his second. The Yellow Bear of Louisiana. 149 quence, the fact remaining that Griffith's Ursm luteolm was based primarily upon the Louisiana animal figured by Major Ham- ilton Smith; hence his name must hold for the species if it is found distinct from the common black bear of the eastern United States (Ursus americanus}. When engaged upon a revision of the North American bears some time ago I was struck by certain cranial and dental pecu- liarities possessed by five skulls* from Prairie Mer Rouge, More- house parish, Louisiana, which led me to regard the species as very distinct from the two species now commonly recognized as inhabiting the United States, namely, Ursus americanus Pallas and U. horribilis Ord. Owing to the absence of skins of this animal, and the rather scanty material illustrative of several other points concerned in a proper elaboration of the group, publication of the review in question was deferred. The recent appearance of an article by Mr. Arthur Erwin Brown,f superin- tendent of the Zoological Garden at Philadelphia, in which this remarkable bear is in part described, though wrongly referred, makes it desirable to issue a preliminary description of the spe- cies, based on the meager material now in hand. The following description is based wholly on the skulls from Mer Rouge, Lou- isiana, of which No. 1155 may be regarded as the type. Cranial Characters. Skull long and flat; fronto-parietal re- gion depressed ; profile of top of skull (including crest) nearly a straight line; sagittal crest long and high, about half the length of upper side of skull in old age. Contrasted with old skulls of male black bears from the Adirondacks, in northern New York, the three old male skulls from Mer Rouge. Louisiana, differ uni- formly in the following particulars : They are longer and flatter ; the occipito-sphenoid length J is greater ; the distance from fora- men magnum to plane of front of last upper molar is greater ; the ratio of zygomatic breadth to basilar length is less ; the ratio of postpalatal length to occipito-sphenoid length is considerably greater. * These skulls have been in the United States National Museum many years and some of their peculiarities were mentioned by Baird in his great work on the Mammals of North America in 1857. f Forest and Stream, New York, Dec. 16, 1893, 518-519. j Occipito-sphenoid length distance from anterior lip of foramen mag- num to suture between basisphenoid and presphenoid. $ Post-palatal length distance from anterior lip of foramen magnum to post-palatal notch, 150 Merriam The Yellow Bear of Louisiana. The largest of the three old male skulls from Mer Rouge, Louis- iana (No. 1155 United States National Museum), affords the following measurements : Basal length (basion to front of pre- maxilla), 292; basilar length of Hensel, 288; zygomatic breadth, 187 ; occipito-sphenoid length, 89 ; postpalatal length, 134 ; dis- tance from inferior lip of foramen magnuin to plane of front of last upper molar, 193 ; interorbital breadth, 68 ; distance betwean postorbital processes, 97 ; occipito-nasal length, 276 ; greatest length of skull, 326. Dental Characters. Molars larger than in any known species of the black bear group (subgenus Eaarctos Gray) ; last upper molar in particular very large and notable for its great breadth as well as length, measuring 30 by 17 mm. in an old male from Prairie Mer Rouge (No. 1155), and doubtless larger in early life, as the tooth is much worn ; the first upper molar in the same specimen measures 19.5 by 15.7 mm. The fourth lower pre- molar is trituberculate, having distinct cusps on the cingulum both anteriorly and posteriorly.* The latter is notched in the middle longitudinally, giving it a double crown. In addition to these cusps, one of the females with less worn teeth than the others has a small but distinct peg-like projection rising from the cingulum on the inner side near the middle, and closely pressed against the main ' cusp, from which it projects only slightly. But this tooth is subject to so much individual varia- tion in bears from the same locality that it would be unsafe to place any reliance on the peculiarity here described unless it is found to hold good in a larger number of individuals than are now available for comparison. Traces of it exist, however, in the other female from Prairie Mer Rouge (No. 988), which is older and has the teeth more worn. Color. The name ' yellow bear ' given to this species by Shaw and Griffith points to a marked peculiarity of coloration, and Mr. Arthur Erwin Brown, in his interesting article already referred to, describes one of his specimens as " flaxen color, with traces of a darker shade on the nape." The skull and teeth of this specimen are not described, but the inference is that they agree with the Ozark skull. Another bear, believed by Mr. Brown to be this species, is described as follows : " The color of * Black bears from the Adirondacks and various other places often have a distinctly trituberculate crown to the tooth in question (pm ? ), but they lack the other peculiarities mentioned. The Yellow Bear of Louisiana. 151 the specimen now living [in the Philadelphia Zoological Garden] is in autumn a rich reddish brown, almost bay. As his coat becomes worn and faded he becomes pale yellowish brown, the color being generally uniform over the body."* This bear was presented to the garden by the late General James S. Brisbin, then stationed at Omaha, Nebraska, from which cir- cumstance Mr. Brown infers that it came from the Rocky Mountain region an inference that hardly seems safe, particu- larly if the animal really proves to be U. htieolus. On the other hand, it is by no means certain that luteolus is always yellow ; and if I were to hazard a conjecture, in view of what little is known on the subject, it would be to the effect that the normal color is black. Geographic Distribution. Very little is known of the geographic distribution of this bear further than the fact that it inhabits Louisiana. It may be found to range over much of the low- lands of the Gulf and South Atlantic states, and to intergrade with the black bear of the mountains of Tennessee and the Caro- linas. A semi-fossil skull from the bed of an old stream near Fort Worth, Texas, examined by me, and the skull found by Professor Cope in a cave in the Ozark hills, in southern Missouri, recently described by Mr. Arthur E. Brown, may belong to an ancestor of this species rather than the species itself. The name cinnainomum of Audubon and Bachmanf cannot be applied to this species, because luteolus has thirty-three years' priority, and also because cinnanwmum was based on an animal from the northern Rocky mountains, which has small molars, like the common black bear of the northeastern United States. *Forest and Stream, December 16, 1893, 518. t Ursus americanus var. cinnanwmum Audubon and Bachman, Quadru- peds of North America, vol. iii. 1854, 125-127. ALPHABETICAL INDEX. A. PAGE Acanthoceras (?) justinaB 17, 38 Adeorbis 101 Alucita kellicottii 52 Ammonites (Bachiceras) pedernalis 102 (Schloenbachia) acute-carinatus. . .102 Ammospermophilus 129, 130, 131 Anazyga 78 Ancylobrachia 72, 73, 77 Anomia texana 16, 22 Araucarites 40 Arctomys flaviventer 138 Arthropoda 17, 39 Arvicola emmonsii 60 operarius 139 riparius borealis 139 Ashmead, Wm. H., elected xii Athyridas 76 Atretia 81 Atrypa 76 reticularis 77 Atrypidae 76, 77 B. Bailey, Vernon, The Burrows of Five- toed Kangaroo Rats vii Baker, Frank, Recent Discoveries in the Nervous System vii Remarks iv, v, vii, xiii Bangs, Outram, elected vii Barbatia parv amissouriensis 16, 26 Bassaris 89 Bean, T. H., remarks iv, xiii Bear, Yellow 147, 150 Beecher, Charles E., Development of the Brachial Supports in Die- lasma and Zygospira. By Charles E Beecher and Charles Schuchert 71 Blodgett, F. H., Notes on the Develop- ment of the Bulb of the Adders' Tongue xiii Brachial supports 71 Brown, Herbert, elected vii Buccinopsis ( ?) parry i 17, 33 Burgess, E. S., xi, xiii Burroughs, J., viii C. Calliostoma 101 Camarella 80 bisculata 80 Caprina limestone 97, 98, 100, 104 Cardium (?) sevierese 16, 29 Centronella 72, 73, 75, 81 Cerithium austinensis 102 obliterato-granosum 102 Certhia familaris americana 63 Chamidas 100, 101, 102, 103 Chelonia 17 Chickering, J. W., The Botanical Land- scape ix Chione (?) decepta 16, 27 PAGE Chrysostoma 101 Cistella 81 Gladophyllia f urcifera ioi Coelosmilia americana 101 Coelospira 76\ 77 barrandei '.'. . .'.77 marginalis 77 Coleopterous larvas 52 Comanche series 9 97 Cook, O. F., Notes on the Natural His- tory of Liberia ix Coralliochama 100 Corbicula arkansaensis 16, 29 Cottontail, Orizaba 143 Cretaceous formations 97 Coville, Frederick Vernon, Characteris- tics and Adaptations of a Desert Flora ix Juncus marginatus and its vari- eties 121 Remarks iv, vii, viii, x, xi, xii Cricetus myoides 55 Crocodilia 17 Cucullaaa comanchensis .'. . 16,25 gratiota 16,25 terminalis, 16, 25 Cylindrites (?) sp 17, 33 Cypridea texana, 17, 39 D. Dall, Wm. H., New Forms of Fossils from the Old Miocene of the Gulf States xi Remarks, etc iv,vi,vii,viii,xi,xii xiii Dayia 74, 80 Dewey, L. H., Geographic Distribution of Grasses in the United States x navicula 74 Diceras (?) 103 Dielasma 71, 72, 73 turgida 71, 73, 77 Dinosauria 17 Diplostoma (?) bulbivorum 113, 114 Dipodomys.. < 83, 85, 90, 95, 96 calif ornicus 91 , 92 merriami 91, 92, 95, 96 phillipii 83,90 phillipsi 83,84,86,87,89,91,92,93,95,96 phillipsii 83,91,95 spectabilis 95 Doran, E. W., Development of the In- testines of Tadpoles viii Doubleday, H. H., elected v E. Echinodermata 16, 21 Epiaster (?) sp 16,21 Eriphyla pikensis 16, 28 Evans, Walter H., elected v Remarks ix Everman, Barton W., The Ichthyologic Features of the Black Hills x Remarks vi, vii, xit 154 Alphabetical Index. F. PAGE Fairchild, D. G., Notes on Apple and Pear Fusicladii v Fernow, B. E., Remarks, etc iv, xii, xiii What are the Special Needs of the Biological Society of Washing- ton xii Foraminifera 16, 20 Fredericksburg division 10, 11 G. Galloway, B. T., xiii Gerboa 89 Gill, T., iv, vii, ix, xi xii. xiii Glauconia (?) 102 Goniolina 40 Goode, G. B., iv, xiii Gurley, R. R., Natural Selection as ex- emplified by the Cackling of Hens ix Gwynia 81 H. Hadena stipata 52 Hallina 72, 74, 78, 82 nifeoletti 74 saffordi 74 Hallock, Charles, The Geographical Dis- tribution of the Musk-Ox vi Hammond, W. A vii Hasbrouck, E. M., The Breeding of the Bald Eagle near Mt. Vernon viii The Development of the Append- ages of the Cedar Wax-wing vii Helicocryp1,us 101 Helicopegmata 73, 76, 77 Hesperomys campestris 60 gracilis 61 leucopus 57, 58 maniculatus 60 myoides 55, 59, 62 Hesperoyucca 42 Hill, Robert T., Notes on the Paleon- tology of the Comanche Series v Paleontology of the Cretaceous Formations of Texas. The In- vertebrate Paleontology of the Trinity Division 9 The Paleontology of the Creta- ceous Formations of Texas. The Invertebrate Fossils of the Cap- rina Limestone Beds 97 Remarks vi Hippurites 103. 104, 108 flabellata 104 texanus 108 Holectypus 101 Holm, T. , iv Howard, L. O., iv, v, viii, xi, xiii I. Ichthyosarcolithus anguis 102, 103 (Caprina) crassipira. . . 103 (?)guadalupse 103 (?) occidentals 103 (?)planata 103 (?)texana 103 Ictidomys 132 Insects, Yucca 41, 47 Isocardia (?) medialis 16, 31 J- PAGE Jackson. Sheldon, The Introduction of Reindeer in Alaska vii Remarks vi Jumping mouse 1 Juncus aristatus 123, 125. 126 aristulatus 123. 125. 12(3 biflorus 123. 126 canaliculatus 123 heteranthos 123, 126 cylindricus 121. 125 marginatus 121. 125. 126. 127. 128 aristulatus 123. 125. 12(>. 127. 128 biflorus 123 paucicapitatus 125 setosus 126, 127 vulgaris 121. 125 odoratus 123 setosus 124. 126 Juvavella... 72 K. Kangaroo rat..... 83 " Mexican 83 Knowlton, F. H iv, xiii L. Lagomys . . 1 17. 138 Lagomys collaris 117. 118. 119. 120 princeps 117. 119, 120 schisticeps 1 19. 120 Leda (?) harveyi 15. 25 Lepidoptera 52 Lepidopterous larvse 52 Lepidotus 17 Lepus arizonae 144 orizabas 143 Lucas, F. A., The Food of Humming Birds vii Remarks, etc iv, viii, ix. xiii Luke, D. D., elected v M. McCormick, L. M., A Hybrid between Pyranga rubra and erythrome- laa viii Macandrevia 72 Macrocolus halticus 91 , 94, 95, 96 Mazama montana 118 Meriones labradorius 2 Merriam. C. Hart, Biology in our Col- leges xi Descriptions of eight new Ground Squirrels of the genera Spermo- philus and Tamias. from Califor- nia. Texas, and Mexico 129 The Fauna and Flora of Eastern Wyoming xii The Four-toed Kangaroo Rats vi Preliminary descriptions of four new mammals from Southern Mexico, collected by E. W. Nel- son 143 Re-discovery of the Mexican Kan- garoo Rat, Dipodomys phillipsi Gray. With field notes by E. W. Nelson 83 Remarks on the Genus Geomys xii Two new Wood rats from the plateau region of Arizona (Neo- toma pinetorum and N. arizpnae). With remarks on the validity of Genus Teonoma of Gray 109 ir>r, The Yellow Hear of Louisiana. Ursus luteolus Griffith 14T Kemurks.etc'.. iv.vi.vii.viii.ix.xi. xii. xiii and Vernon Bailey. Notes on Biological Reconnoisance of Wyoming xi Merriam. Florence A. . elected vii Miller. Jr.. Gerrit S.. A New Jumping Mouse (Zapiix inxiqnix) vi A Jumping mouse (Zaputt inxir/nin Miller) new to the United States .... 1 Description of a new White-footed mouse from the f^asteru United States 55 Notes on Thomomys bulbivorus. ... 1 13 Modiola branneri 16. 24 Mollusca 22 Molluscoidea 16. 22 Monopleura marcida 1(5. 29. 102. 103 pinguiscula 16. 29. 102. 103 subtriquetra 103 texana 103 Moore. V. A.. The Distribution of Patho- genic Bacteria -in the Upper Air Passages of Domesticated Ani- mals x Observations on the Distributions and Specific Character of the Streptococci Group of Bacteria ix See also Smith. T.. and Moore. Mus agrarius americanus 59 bairdii 61 . 62 michiganensis 61 noveboracensis 59 sylvaticus 59 noveboracensis 59 Musculus leucopus 59 Mynomes ... 139 N. Natica (?) texana II Natica ( Amauropsis) avellana Nelson. E. W.. A New Species of Lago- mys from Alaska Description of a new Species of Lagomys from Alaska : Description of a new Species of Arvicola. of the Mynomes group. from Alaska (See also Merriam.) Nelson's. Henry Clay, death Neotoma arixonae 109. 1 10. cinerea 110. fuscipes 109. mexicana 111. pinetorum 109. Nerinea austinensis 1 7. 36. cultrispira subula Neritina sp r, Neumayria walcotti 11 Newberria Nordqvist. Dr.. a guest Remarks Nucleatula Nucleospiridtp 5.32 .'102 viii .109 111 112 111 112 111 101 .101 .102 '. 37 '.37 o. Obolella. 80 Officers elected Dec. 31. 1892 iv Officers elected Dec. 30. 1893 xiii Orbitulites texana 40 Ostrea franklini... ...16. 23 joana muusoni... Otospermophilus . Ovis canadensis. . . dalli .... PACJK 105 102. 105 133 13H ..118 P. Palmer. T. S i v . x , xiii Parasmilia austinensis 101 Patella 101 Patellina texana J6. 20 Paterina so Pecten stantoni 16. 24 Pedregal 87 Pentamerus 80 Perognathus 85. 89. 95, 96 Perodipus 89 Pholadomya knowltoni 1 6. 30 lerchi 16. 30 Pileqlus 101 Plagioptychus (?) cordatus 103 Plantae 17. 39 Pleurocora coalescens 101 texana 101 Pleuromya (?) henselli 16. 31 Pollination. Yucca 41 Pratt. F. C.. elected xii Prodoxids 50 51 53 Prodoxus 47. 52 cinereus 48. 51 coloradensis 47 decipiens 49 intermedius 49 intricatus 50 reticulatus 48, 50 Pronuba 41. 43. 52. 53 Pronuba maculata 41. 43. 44. 45. 46. 51. 53 synthetica 46. 53 yuccasella 42. 44. 45. 46. 49. 52 Protocardia sp -.16. 29 Pterophorid 52 Pterophoridae 52 R. Radiolites 102. 104. 108 Radiolites austinensis 100. 103 davidsonii 102. 103. 106. 1(W texana 102. 103 Rathbun. R iv. xiii Rensselaeria . 72 Requiena bicornis 103 patagiata 102. 103 texana 102. 103 texana (?) 16.28 Reyburn. R viii Riley. C. V.. Further notes on the Yucca Insects and Yucca Pollination 41 Further Notes on Yucca Polli- nation xi Remarks, etc iv. viii. x.xi. xii. xiii. Rives. William C.. elected viii Rose. J. N.. Two Trees of Economic Im- portance from Mexico ix Remarks . xii Rostracea 80. 81 Roth. Filibert. elected xiii Rudistae... 103 Rudistes... ...99.100.102 S. Schaetter. E. M xi Schuchert. Charles, elected x Development of the Brachial Sup- 156 I'ACK ports in Dielasmu and Zygospira. By Charles E. Beecher and Charles Schuchert. And on the Development of the shell of Zygospira recurvirostra. By Charles Schuchert 71 Remarks iv Seaman. W. H xiii Simpson. Charles Torrey , elected xi Sitomys 55 americanus . .55.56.57.59.60.62.63.64.65.66 arcticus 67 aureolus 62 canadensis.55.56.57.59.eO.61.62.63.67.6K.69 Smith. Erwin F.. Additional Notes on Peach Rosette v A Bacterial Disease Cucumbers. etc.. working through the Fib- rovascular Bundles and prob- ably transmitted by Insects xiii Peach Yellows and Plant Nutrition . . x Remarks iv. v. ix. x The Symbiosis of Stock and Graft. . .x Smith. Theobald. The Bacteriology ( of Potomac Water and its Bearing upon Sanitary Problems viii A New Sporozoon in the Intestinal, Villi of the Ox "... x Remarks iv, ix, x. xm and V. A. Moore. The Growth of '. Bacteria Through the Pasteur- Chamberland Filter v Solarium planorbis 10:2 Spermophilus annectens 133 beecheyi 133 beecheyi flsheri 133 chrysodeirus 134 brevicaudus . . 134 elegans 131 mexicanus 131. 132 . nelsoni . . .. 129 obsoletus '..].'. 132 perotensis 131 . 132 spilosoma 131 " annectens 132 major 131 Sphan-ulites 108 SpiriferidfE 76 Stanton. T. W iv. v, yi Stejneger. L yii Sternberg. Geo. M. . elected xiii Stiles. C. W.. Artificial Species of Ces- todes xii The Cause of Measly Duck ix Sudworth. G. B iv Swingle. W. T.. Some Problems of Plant Geography i n Florida xi T. Tamias alpinus 137 callipeplus 136 minimus pictus 137. 138 panamintus 134. 135 speciosus 137 Teonoma 109. 1 12 Terebratellidae 72 Terebratulina 81 Thomomys 113 bottae 115. 116 bulbivorus 113. 114. 115 Mt. Orizaba.... 145 orizabse 1 45. 146 peregrinus 145. 149 wandering 146 Thompson. J. B. . elected xii Townsend. C. H.. The Propagation of the Atlantic Coast Oyster on the PACK Pacific Coast vi Sea-Otter Fishing in Alaska in Re- lation to the Natives yi Remarks % vii Trigonia crenulata! 16. 27 stolleyi 16,26 Trinity division 11. 12. 17 fossils .....15 Trochus texanus 102 Troglodytes hiemalis 63 True, F. W iv. xiii Trullacea ' , 80. 81 Tylostoma pedernalis 17. 33 U. Ursus americanus 147. 149 cinnamomeus 147 horribilis _,..'.... Hi* luteolus ... ... 147. 148. 149. 151 Van Deman. H. E., ix Vanessa cardui . . . 51 Vasey's. Geo-rge. death .; viii Vermes ;>,. .<. , .x 16. 21 Vertebrata .-li.-.-^ 17 Vicarya branneri 17. 33 Viviparus (Natica ';) cossatotensis 17. 32 A.< W> Waite. M. B.. The Destruction of Lich- ens on Pear Trees vi The Variation of the Fruit of the Pear due to Difference of Pollen. . .vii Remarks v. viii Walcott. C. D iv. viii. xiii Waldheimia 72. 80 bicarinata 74 mawii 74 Ward. L. F.. The Flora of the Trinity Division of the Comanche Series in Texas vi The New Botany x Weismann's Concessions xi Remarks, etc iv, v. vii. viii. xi. xiii Washita division 10. 11 Xerospei'mophilu!- X. Y. 131 Yucca 41. 53 aloifolia 43.47.53 angustifolia glauca n3 australis 53 baccata 45. 53 brevifolia 3 elata 45. 53 filamentosa 53 filifera ?? gloriosa - guatemalensis 53 insects 41. 47 pollination .-*1 rupicola 45, 53 ti-eculiana 53 whipplei 41 .42. 43. 45. 48. 52. 53 graminit'olia . 45. 47. 51 . 1 lphabeti