Q 11 .U563 MSRLSI vi IN A PSAtE OK aE Ny Pr aay = PLATE 1 BULLETIN 171 U. S. NATIONAL MUSEUM yy) Zp Topographic map showing location of Cumberland Cave and its relation to the physiography. The cave is indicated (X) on north end of spur just south of Corriganville, Md. Map is Scale, 1/62500, or about taken from U. S. Geological Survey Frostburg Quadrangle. Contour interval, 20 feet. 1 mile to the inch. SMITHSONIAN INSTITUTION UNITED STATES NATIONAL MUSEUM BuLuetin 171 THE PLEISTOCENE VERTEBRATE FAUNA FROM CUMBERLAND CAVE MARYLAND BY JAMES W. GIDLEY anp C. LEWIS GAZIN UNITED STATES GOVERNMENT PRINTING OFFICE WASHINGTON : 1938 LSS La ee For sale by the Superiatendent of Documents, Washington, D.C. - - - - <= - - Price 25 cents (paper) ADVERTISEMENT The scientific publications of the National Museum include two series, known, respectively, as Proceedings and Bulletin. The Proceedings series, begun in 1878, is intended primarily as a medium for the publication of original papers, based on the collec- tions of the National Museum, that set forth newly acquired facts in biology, anthropology, and geology, with descriptions of new forms and revisions of limited groups. Copies of each paper, in pamphlet form, are distributed as published to libraries and scientific organizations and to specialists and others interested in the different subjects. The dates at which these separate papers are published are recorded in the table of contents of each of the volumes. The Bulletin series, the first of which was issued in 1875, consists of separate publications comprising monographs of large zoological groups and other general systematic treatises (occasionally in several volumes), faunal works, reports of expeditions, catalogs of type speci- mens and special collections, and other material of similar nature. The majority of the volumes are octavo in size, but a quarto size has been adopted in a few instances in which large plates were regarded as indispensable. In the Bulletin series appear volumes under the heading Contributions from the United States National Herbarium, in octavo form, published by the National Museum since 1902, which contain papers relating to the botanical collections of the Museum. The present volume forms No. 171 of the Bulletin series. ALEXANDER WETMORE, Assistant Secretary, Smithsonian Institution. WasHINGTON, D. C., January 26, 1988. Ir Se — Z\ASONIA, ZAM 7 Ay i) A ( FEB CONTENTS Matroductione. oes oo a> oe LEE SESS Sea ee Sea en oe History of investigation....--.------+---~------+-+<22----2+-----=--- Location of Cumberland Cave Accumulation and preservation of migterial] os Oe ee Relationship of cave to physio graphy... ----------------------------- Gumberland Cave fauna.--_-----=++-------~-----=-45-+-=+2---+-=--- Environment of the fauna -------------------------------------- Age and relationships of t Heal Ase ee i eS aa aa Systematic description of vertebrate fauna__._-2_--------+-----=------ Order Loricata. 4.2222 2221-22222 242-2 2-2-2 2 2 - eee nnennae = Grocadwaid tees oi ae 22s as See nee eee Order Serpentes.=+-----------2425-+--42-22--25- 25-5 eee noone Ophidian remains------.-------------------=---------------- @rder-Galliformes= 2+ ---.--22 22242. =. 2222-2 2 = eee enna Family Tetraonidae-_--------------------------------------- Bonasa umbellus (Linmseus) 224 222222. Sa === = @rderinsectivoralt= 22 2.==52 == 222225--2-2--=------ == nae amily Soricidae=++2-++=-2--+2=+----==+-----2---=---=""-=- ores Spuse=sh22-5-¢5-chhe2 set essene Sse e-- 6 Sana anee Blarina brevicauda (Say) -------------------------------- @ider @hiraptera- 2-22 S222 22-2 2S Sele ae == Siete a Family Vespertilionidae------------------------------------ Eptesicus cf. grandis (Brown) 22 22-282) RU seo-- === Corynorhinus alleganiensis Gidley and! Garin? Seee so te ee Grder Garmivotal 2-22: 35-2429 =2225-552522242=2=--~--- === Family Canidaessss22: $2222 2222222-222+22255~---=-- 22S oe Canis armbruster feGidlay sa fusees -Satlsks 4S SoSe Ses Canis ef. priscolatrans Cope-.---------------------------- BamilysUreidaess) 22522 F242 23 22c Baas sash == 4 —-- 2 oo e Euarctos vitabilis Arctodus haplodo Family Mustelidae - - Martes parapenn Mustela cf. vison Gulo gidleyt Hall (Gidley) i242 222 shs22se 235 ~ = 22 = == = fh (Cope) Be Saas eee ea a= oe anti Gidley and Gazin- sy Ne \ sR Sehirebers= se ee 2 ee Pe eae ase Lutra parvicuspis Gidley and Gazin_--------------------- Brachyprotoma p ristina Brown: 2222222222252 22=-2-=-=---=- Spilogale marylandensis Gidley anid) Cazine =) Sewers ee Se Taxidea marylandica Gidley and Gazin------------------- anilysWelidwe>\] sss DUNE rears ore en ee aa Felis cf. inexpectata (Cope) ------------------------------ Foliemear atrov Leidy 22 2! = 2222 2522252 s24==-3422--2-2-== IV BULLETIN 171, UNITED STATES NATIONAL MUSEUM Systematic description of vertebrate fauna—Continued. Page Order: Rodentisia=2. 232 = ee ae ee ee ee ee eee 53 Hamily Sciuridaes." 2. oa eae ee ee ee ee 53 Marmotavmnonacc) (ui seus) arse: aye er eee eae 53 Citellus ef. tridecemlineatus (Mitchill)......._....._________ 54 Tamias cl. striatus: (hinnaens) eee see ee ee 54 Sciurus (Tamiasciurus) tenwidens Hay__._____.__._______- 54 Glaucomys: spe ae ee Se ae Sl a See ee ee ee 56 Family -Geoomyidae.s 3 = ese a ee ee 57 Plesivithomomys potomacensis Gidley and Gazin__-__________ 57 Family" Castoridae. =: sso 2ees 2222 5. ee eee eee 59 Castonm canadensis Kohl a a= ee ee 59 Family ‘Cricetidae. 22 2 S22 2s. ee 59 Peromyscus cf. lewcopus (Rafinesque)...._._......._--_-_._ 59 Neotomasmagester: Dard eee es oe ee ee 59 Parahodomys spelaeus Gidley and Gazin.-_._.._..__--_.-- 60 Synaplomys cl. coo pen Dard. anes a nae oe ee 62 SUNG DLOMYS MACON YS) Spe ae ee ee 63 Microtus (or Pitymys?) cf. involutus (Cope). ___---._.------ 63 Ondatrarctannectens: (Brown) 29a" soe ae ee 65 Family, Zapodidacs— 222 so. es Se 67 LODUS BD ea ee ee ae ree eee, Seah a 67 Napacozapusch. tasrgnis. (Miller) S22 7 9. Se ee eee 67 Kamiuly Erethizomtidaes 2 32.2 ea ee 67 Hrethizon cl. dorsatum.(einnaeus)) == 3 ee ae 67 Order Lagomorpha _-______------ Cert ee i: ee Bak anes ria 68 Family Ochotonidae_______- wc oN) eS SD ele ne pe ee 68 OCTIOLON GUST) Baars te te Se rn ee ey or eee 68 Family Leporidae: _... =... et Saxe ipl he Os ee glee eee 69 Lepus cf. americanus Mrxiebens. = 2.2. = enon a eee 69 Order Proboseides: . 6. 23-22. Saale es seen oy ee oe 70 Family Mastodontidae so: )\ 422 ae eee 2 eee ee ee 70 Mamimutsck.. americanwin: (Nernst 70 Order Perissodactyla 224 8 2 ae ee ee ee es 70 WamnilyBiguidae.s. == ot ete ee oes ee get een a 70 FguusiSD. osc ce coe ee 32 a a ee see 70 Family Tapiridaé.. #s2“25.. 2 22-2 eee ee eS 70 TODiTUS Spi. ee oe ee A ee ee ee 70 Order Artiodactyla: 2.282232. ee = So a ee 71 Vamily Tay assuidae 3 soe. 5) 2 a ee 2 71 Platygonus cumberlandéensis Gidley. -22-— ee tee 71 Platygonus vets (2) eid 5) 532 ae ee ee ee 82 Malohyus: exortivus Gidléy=— 22 5a seer See eee 83 Mylohyus cf. pennsylvanicus (Leidy)__........--.-------- 86 Family ‘Cervidae 23 <2 oe scat be eee ee ee 86 Cervus'S Pies 2nd 2 See eS ee ee eee 86 Odocoileus cf. virginianus (Boddaert)__.._.._.__._._-.-------- 86 Family, Bovidsaeteeeu2- 52 S642 2 ee ee ee ee ee 86 Euceratherium(?) americanum (Gidley) _._..-...---.------ 86 Literature cited: 2. 224.45 eh ee eee re ee 91 MARS ae Sr So nye ee Are, nT ec a 95 —_ 16. he 18. Ls 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. ILLUSTRATIONS FIGURES _ Skull and upper dentition of Corynorhinus alleganiensis Gidley and WSkullbol Canisianmorustert: Gidley 2s 45 = eas ee eee eee eee . Skull and lower carnassial of Canis armbrusteri Gidley _------------- Per cunt tak Oras Mc TiT7a Nee SC Ty Gx IL a pen eee Parle ot Ga rau sai L0 Tess Cerium Onl CU Cys eee ee ae . Left ramus of mandible of Canis armbrusteri Gidley -_-------------- . Right ramus of mandible of Canis armbrustert Gidley--------------- . Left ramus of mandible of Canis armbrusteri Gidley __-------------- . Right ramus of mandible of Canis armbrusteri Gidley----~---------- . Right ramus of mandible of Canis armbrustert Gidleysn nse ae we . Right ramus of mandible of Euarctos vitabilis (Gidley) -------------- Skull and mandible of Fuarctos vitabilis (Gidley) ---~---------------- PASiculOt archos UILQUULUS, (GiGIGy a2 oe epee ee PaSkullof uanrchos vraurits. (GiGlCY 222 ae ce ee ee ee ee . Anterior portion of male skull, right rami of female mandibles, and left ramus of male mandible of Martes parapennanti Gidley and Gazin_- Anterior portion of skull and right ramus of male mandible of M ustela GIA i STET SCHROMEIS Best aa ea eee a ate Beer Aree Seen Anterior portion of skull and mandible of Mustela ef. vison Schreber_- Right ramus of female mandible of M/ustela cf. vison Schreber-------- Skull ane menditsteof Gauloigzuareys tall 2-2) 2 Slaull"@f Grlovgidlere: tial Sa St ea A a eee Maxillary portion of skull of Lutra parvicuspis Gidley and Gazin_--- Skull of Brachyprotoma pristina Brown__-_--_----------------------- Left ramus of male mandible of Brachyprotoma pristina Brown_------ Right raraus of mandible of Brachyprotoma pristina Brown- --------- Right ramus of mandible of Spilogale marylandensis Gidley and Gazin_ - Skull and mandible of Taxidea marylandica Gidley and Gazin__------ Skull of Taxidea marylandica Gidley and Gazin_----~--~------------- Upper carnassial and left ramus of mandible of Felis cf. inexpectata (SONG) ee ea ee ee pple rgd Noe erg SL Ped Hoot DOMES OL MELTS MEAT GETOT Ln Cyt ames ee eee ere Skull and mandible of Sciurus (Tamiasciurus) tenwidens Hay _~------ Left ramus of mandible of Plesioithomomys potomacensis Gidley and Left ramus of mandible of Neotema magister Baird. __-------------- Right ramus of mandible of Parahedomys spelacus Gidley and Gazin_- Right ramus of mardible of Parahodomys spelaeus Gidley and Gazin_ Right ramus of mandible of Microtus (or Pitymys?) ef. involutus (Cope) - Left ramus of mandible of Ondatra cf. annectens (Brown) ------------ Skull and mandible of Erethizon cf. dorsatum (Linnaeus) - - - --------- Rastilower molarot.laperus spot -) 522 ee ee ee eee ea Skull of Platygonus cumberlandensis Gidley ------------------------ Page 14 18 19 21 21 22 22 22 23 28 29 30 31 34 36 37 37 38 39 42 43 44 44 46 48 50 52 55 58 60 61 61 64 66 68 a 73 vI BULLETIN 171, UNITED STATES NATIONAL MUSEUM Page 40. Skull of Platygonus cumberlandensis Gidley_.._.--.---.------------ 74 41. Skull of Platygonus cumberlandensis Gidley.....-.--.-------------- 75 42. Skull of Platygonus cumberlandensis Gidley_..-..---.-.------------ 76 43. Skull of Platygonus cumberlandensis Gidley_..-.-.---------------- a 44. Skull of Platygonus cumberlandensis Gidley----..-.-.-------------- 78 45. Left hind foot of Platygonus cumberlandensis Gidley __....---.------ 79 46. Right ramus of mandible and cheek teeth of Mylohyus exortivus Gidley - 83 47. Cheek teeth of Mylohyus exortivus Gidley__..-.....-..---.--..---- 84 48. Superior deciduous premolars and first permanent molar of right side of Mylohyus exortious Gidley 2...) 32 Se 84 49. Superior cheek teeth of Euceratherium(?) americanum (Gidley) -__-_-- 87 50. Left ramus of mandible and right lower premolars of Euceratherium(?) americanum (Gidley) = 32626222 a id Be eee 88 PLATES Facing page 1. Topographic map showing location of Cumberland Cave and its rela- tion tothe pliysiopraphy.2_- <= ee ee ee ee ee Frontispiece 2. Cut made by Western Maryland Railroad Co., which uncovered the Camberland Cave deposit 22. 3.22222 os.52 cae ee eee 3. View from opposite side of railroad cut, showing fossil deposit at bot- tom, near track, and traces of ancient opening at top of cliff_____-- 3 4. Near view of part of excavation made in fossil deposit___.._.__.-__-- 6 5. Later view of opening to excavation made in fossil deposit. _______-__ ti 6. Small portion of cave interior, showing bones sticking in clay and stalactitic material. 2-23.52 So cee oe ee er 10 7. View from spur in which cave is located, looking north across Wills Greek Valley. 2c ce ae ay hg ee a oe ee ee Bl 8. Mounted skeleton of bear (Euarctos vitabilis Gidley), from Cumber- land Cave, exhibited in U. S. National Museum_____-___-------- 26 9. Mounted skeleton of wolverine (Gulo gidleyi Hall) from Cumberland Cave, exhibited in U. S. National Museum________._____.__----- 27 10. Mounted skeleton of peccary (Platygonus cumberlandensis Gidley) from Cumberland Cave, exhibited in U.S. National Museum- - - ---- 70 THE PLEISTOCENE VERTEBRATE FAUNA FROM CUMBERLAND CAVE, MARYLAND By JAMES W. GIDLEY and C. LEWIS GAZIN INTRODUCTION LimMESTONE caverns have played a major role in preserving a record of the Pleistocene life in the eastern region of North America. Except for the remarkable deposits in Florida, our knowledge of the com- position of Pleistocene mammalian faunas from this region has been limited largely to occasional and fortunate cave finds. This mode of accumulation rivals that of the tar pits in presenting a broad cross. section of the land life belonging to the geclogic period immediately preceding the present. Caves are abundant in the limestone areas of the Appalachian and Mississippi Valley regions, but very little systematic exploration has been done with a view toward obtaining representative collections of Pleistocene mammals. Strong impetus for exploratory work has come from the field of archeological research, and many of the important paleontological finds were made during the investigations for the remains of man. The search for saltpeter has also led to the exam- ination of many caves, but in this work the importance of fossil re- mains found has undoubtedly been overlooked in the majority of instances. Cumberland Cave was first found by workmen in excavating for a railroad cut near Cumberland, Md. Considerable fossil material was destroyed by steam shovel and dynamite before the significance of the find was known, but the abundant material subsequently obtained includes a remarkable variety of mammalian forms, many of which are represented by unusually well preserved remains. The junior author is greatly indebted to C. W. Gilmore, curator of vertebrate paleontology at the National Museum, for advice and criticism during the preparation of this report, and to Dr. J. B. Reeside, Jr., and Dr. W. C. Alden, of the United States Geological Survey, for a better understanding of the geologic and physiographic relations of the cave. Acknowledgment is also made of the courtesy extended by G. S. Miller, Jr., Dr. Remington Kellogg, and staff of the division of mammals in the National Museum and Dr. H. H. T. Jackson, EK. A. Goldman, and A. H. Howell, of the United States Biological Survey, in giving generously of their time and advice and 1 2 BULLETIN 171, UNITED STATES NATIONAL MUSEUM in permitting unreserved use of the collections of recent mammals, Dr. Walter Granger, of the American Museum of Natural History, and Dr. C. L. Camp and R. A. Stirton, of the University of California, have aided greatly in loaning type material from their cave collections. The drawings included with the text were made by Sydney Prentice and Rudolf Weber. HISTORY OF INVESTIGATION The attention of the United States National Museum was directed to the occurrence in 1912 by Raymond Armbruster, of Cumberland, Md., and by George Roeder, of Swetnan, Va. Investigation of the occurrence was undertaken by Dr. J. W. Gidley, and quarrying operations in the cave were conducted by him at intervals during the years 1912 to 1915. For a part of this time he was assisted in the excavation work by Mr. Armbruster. The exploration work carried on at the cave is interestingly described in three popular papers published in “Explorations and Field-Work of the Smithsonian Institution” for the years 1913, 1914, and 1915. Early results of Gidley’s study of the fauna were published in two short papers that appeared in 1913. These included a description of an “eland’”’ (1913a) (see also Gazin, 1933), and new species of wolf and black bear (1913b). In later papers Gidley (1920a, 1920b) gave a general account of the occurrence and described the peccary remains taken from the cave. Dr. Alexander Wetmore (1927) reported the remains of a ruffed grouse included in the collection. Since the death of Dr. Gidley in 1931 it has become the junior author’s privilege to continue the investigation of the Cumberland Cave fauna, completing the manuscript and describing those portions of the fauna that had not been studied. In 1933 a revised list of the forms encountered in the investigation was published (Gidley and Gazin, 1933) together with descriptions of the remaining new forms in the fauna. LOCATION OF CUMBERLAND CAVE Cumberland Cave is located in Allegany County, western Maryland, about 4 miles northwest of Cumberland (see pl. 1). It is situated in a cut made by the Western Maryland Railroad through the north end of a ridge just south of Corriganville, near where Jennings Run joins Wills Creek. Its location as determined from the topography shown on the Frostburg quadrangle, U. S. Geological Survey, is at about latitude 39°41’ N. and longitude 78°47’ W. The cave is exposed in the cut at the base of an escarpment of Devonian limestone about 75 feet high. Dr. J. B. Reeside, Jr., of the United States Geological Survey, informs the writer that the PEATE, 2 BULLETIN 171 U. S. NATIONAL MUSEUM ‘yisodap JAV) puevylequin’) oy potaAooun Yor M “40{5) PVOIT[IL | puvyAlryy uJ So M Aq ope w yng PLATE 3 BULLETIN 171 U. S. NATIONAL MUSEUM near track, and 9 at bottom deposit ssil fo Photographed in 1913 by Raymond Armbruster. showing ilroad cut Tal f yosite side o traces of ancient opening at top of cliff. 1e@W fre mM OPT Vi PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 3 limestone at this point belongs to the Keyser member of the Helderberg formation, and is about 50 feet stratigraphically below the base of the Oriskany. The limestone is structurally a part of the steeply dipping west flank of the Wills Mountain anticline. In the vicinity of the cave the strata are standing nearly vertical, a fact that probably aided greatly in the penetration of surface waters. ACCUMULATION AND PRESERVATION OF MATERIAL The following account of the conditions encountered at the cave is taken from Gidley’s preliminary report (1913b, pp. 94-95) on the occurrence: In making the railway cut, several small chambers at bigher levels than the one containing the bones were encountered, and before the work of excavation began there was said to have been an opening to the surface on the crest of the hill directly above the middle of the present roadbed. This opening so nearly overhead probably at one time served as a trap through which were introduced the animals whose remains are now in the deposits of the bone cavern. ‘There are other openings along the line of outcrop of the ledge, one of them at about the same level with the bone-bearing deposits, appearing at the north end of the ridge where it slopes abruptly down into the Wills Creek Valley. These openings may or may not have communicated at one time with the caverns intersected by the railroad cut, but probably had nothing to do with the accumu- lation of material in the latter. From Brown’s [1908, p. 163] account of the Conard Fissure, it would seem that the conditions governing the accumulation of material in the Cumberland Cave were quite similar. The bones for the most part are much broken, yet show no signs of being water worn. They are found scattered fairly uniformly throughout the entire mass of unstratified accumulations which consist entirely of cave clays and breccias, unevenly hardened and more or less cemented together by stalactitic materials. There is an almost entire absence of admixture of sand or gravel, or in fact anything that would suggest the possible aid of stream currents in sorting or placing the material during the process of accumulation. RELATIONSHIP OF CAVE TO PHYSIOGRAPHY Five geographic provinces are recognized in Maryland (Fenneman, 1930): Coastal Plain, Piedmont Plateau, Blue Ridge, Valley and Ridge Province, and Appalachian Plateau. The region about Cumberland Cave belongs physiographically in the western part of the middle section of the Valley and Ridge Province. Physiographic studies of western Maryland and vicinity (Abbe, 1900; Clark and Mathews, 1906; Stose and Swartz, 1912; Stose and Miser, 1922) show that several surfaces, representing periods of ces- sation of downward stream-cutting, have been developed. Rem- nants of the various upland surfaces recognized are commonly represented by sets of ridge tops having accordant levels. More recent stages in the process of stream erosion are seen in the different cravel-strewed terraces. 4 BULLETIN 171, UNITED STATES NATIONAL MUSEUM Perhaps the oldest surface recognized, which has been referred to the Schooley Plain, includes the higher portions of the Allegheny Mountain section of the Appalachian Plateaus. A somewhat lower level may be represented by the crest of Wills Mountain (see pl. 1) just east of Cumberland Cave and having an elevation varying from 1,400 feet to 1,800 feet. A still lower surface is believed to be indi- cated by the crest of Shriver Ridge just north of the town of Cumber- land and by a spur of Knobly Mountain to the south of the town. W. B. Clark and E. B. Mathews (1906, p. 89) correlated these ridge crests with the Harrisburg surface, and Cleveland Abbe, Jr. (1900, p. 52), considered them as representing a surface which he called the Shenandoah Plain. The ridges have an elevation of about 1,100 feet near the Potomac River, and if the surface represented were extended up Wills Creek Valley at the present stream gradient it would have an elevation of about 1,200 feet at Cumberland Cave. It may be represented, as suggested by Clark and Mathews, by the ridge west of Wills Mountain, in which Cumberland Cave is located. This ridge crest varies in elevation from about 1,040 feet near the cave to about 1,300 feet farther south. The elevation of the cave deposit, as determined by O. P. Hay (1923) is 837 feet. Assuming practically no gradient for the surface between Cumberland and Cumberland Cave, we have a minimum difference of nearly 300 feet between the elevation of the cave and that of the surface. During later stream history (Abbe, 1900) two important terraces were developed in the vicinity of Cumberland at elevations of about 800 feet and 650-700 feet. The higher of these may have had an elevation of about 900 feet on the spur in which the cave is located. This is probably very close to the elevation at which the cave mouth was located. The lower terrace may have been 750-800 feet in elevation near the cave. This level appears to be slightly lower than the cave deposit. It seems evident that the fossil material did not accumulate until cutting by Wills Creek had at least reached the stage represented by the higher terrace, and probably at a somewhat later date when the drainage system had dissected this surface. PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 5 CUMBERLAND CAVE FAUNA The classified list that follows includes all the Pleistocene vertebrate fauna identified from the Cumberland Cave and herein treated: Loricatva: Crocodylid. SERPENTES: Ophidian remains. GALLIFORMES: Tetraonidae: Bonasa umbellus (Linnaeus). INSECTIVORA: Soricidae: Soreaz sp. Blarina brevicauda (Say). CHIROPTERA: Vespertilionidae: {Eptesicus cf. grandis (Brown). tCorynorhinus alleganiensis Gidley and Gazin. CARNIVORA: Canidae: Canis armbrusteri Gidley. tCanis cf. priscolatrans Cope. Ursidae: }Huarctos vitabilis (Gidley). tArctodus haplodon (Cope). Mustelidae: +Martes parapennantt Gidley and Gazin. Mustela cf. vison Schreber. tGulo gidleyz Hall. tLutra = parvicuspis Gazin. tBrachyprotoma pristina Brown. +Spilogale marylandensis Gidley and Gazin. }Taxidea marylandica Gidley and r Gazin. Felidae: }Felis cf. inexpectata (Cope). {Felis near atrox Leidy. RODENTIA: Sciuridae: Marmota monaz (Linnaeus). Citellus ef. tridecemlineatus (Mitch- ill). Tamias cf. striatus (Linnaeus). tSciurus tenuidens Hay. Glaucomys sp. Geomyidae: tPlesiothomomys potomacensis Gid- ley and Gazin. Gidley + Extinct species. t Extinct genera. RopeEntT1A (Cont.): Castoridae: Castor canadensis Kuhl. Cricetidae: Peromyscus cf. leucopus (Rafin- esque). {+ Neotoma magister Baird. tParahodemys spelaeus Gidley and Gazin. Synaptomys cf. cooperi Baird. Synaptomys (Mictomys) sp. +Microtus (or Pitymys?) cf. in- volutus (Cope). tOndatra cf. annectens (Brown). Zapodidae: Zapus sp. Napaeozapus ef. insignis (Miller). Erethizontidae: Erethizon ef. dorsatum (Linnaeus). LAGOMQRPHA: Ochotonidae: Ochotona sp. Leporidae: Lepus cf. americanus Erxleben. PROBOSCIDEA: Mastodontidae: {Mammut cf. americanum (Kerr). and | PerrssoDACTYLA: Equidae: {Equus sp. Tapiridae: {Tapirus sp. ARTIODACTYLA: Tayassuidae: tPlatygonus cumberlandensis Gid- ley. tPlatygonus vetus(?) Leidy. t{Mylohyus exorlivus Gidley. tMylohyus ef. pennsylvanicus (Leidy). Cervidae: Cervus sp. Odocoileus ef. virginianus (Bod- daert). Bovidae: tEuceratherium(?) americanum (Gidley). 6 BULLETIN 171, UNITED STATES NATIONAL MUSEUM ENVIRONMENT OF THE FAUNA Evidence of the environmental conditions- that prevailed ir the region of the cave during the time in which the fossil material accumu- lated might reasonably be expected from a consideration of the various forms present in the fauna. Many of the species recognized are closely related to living forms, and it seems probable that their habits were not greatly different from those of living representatives. The fauna from the cave includes a peculiar assemblage of animals. Many of the species are comparable to forms now living in the vicinity of the cave, but others are distinctly northern, or Boreal, in their affinities and some are related to species peculiar to the southern, or Lower Austral, region. The fauna now living in this part of Maryland is composed principally of species representative of the Carolinian or Upper Austral region and of the Alleghenian division of the Transition Zone (Merriam, 1900, p. 292). The wolverine, Gulo, and lemmings of the group Mictomys are distinctly northern in range, and it seems highly improbable that they were contemporaneous in the same region with a crocodylid and a tapir. Such an association might be plausible with sufficient topo- graphic relief, but from a consideration of the physiographic history it seems evident that during the time in which the cave deposit accumulated relief was not greater than at present, as streams may not have cut so deeply below the old upland surfaces. A more probable explanation of the apparent association is that the fauna is not entirely contemporaneous and that the entombment extended over a period of time sufficient to allow important climatic changes to take place. The cave may have received material during a portion of a glacial and of an interglacial stage. Gidley (1920a) contended that the forms were contemporaneous in the same region, postulating greater relief, but the physiographic evidence does not seem to justify his reasoning. Included among the forms that may well have been associated with wolverine and lemming mice in the colder fauna are the long- tailed shrew, fisher, mink, red squirrel, muskrat, porcupine, jumping mice, pika, hare, and elk; although in present distribution several of these extend well into the southern region. Suggesting warmer climatic conditions are bats and peccaries, in addition to crocodylid and tapir. Among those forms whose most closely related living representatives are now found much farther west are coyote, badger, pika, a pumalike cat, and a pocket gopher that resembles Thomomys more closely than it does Geomys. The fauna as a whole is strongly indicative of a wooded region with no lack of moisture. The floristic condition is suggested particularly well by the spermophiles and jumping mice, and the plentiful supply P71 Pe AtiE A BULLETIN U. S. NATIONAL MUSEUM Aq ¢16l ul I payde Ise yy UU *Sutuodo 18 sulpue “To}SNIQu4sy put UAB YT 1S Ad I I D1 ad JIS dap Issoy ul opel uole NCIX Q o fo }1IeC { j MOIA Ie oN FZ ATES BULLE EmINe ait U. S. NATIONAL MUSEUM HIG USE oyd e150] 04] “A 9| I *1oqSNAIquUIy puow Avy] Aq IL) “CZ st oansy sraddq yisodap l I Ss Sof ul 3 I Iv UW UOlle. \v IX9 O} SULUDC { ( ) PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE a of water is evident from the presence of fisher, mink, otter, beaver, bog-lemmings, and muskrat, as well as of the crocodylid. The greater part of the fauna is not unsuited to humid conditions and an abun- dance of vegetation. However, the horse, coyote, badger, and hare are suggestive of a more open terrane, perhaps bordering a wooded area. It should be noted, moreover, that coyotes and perhaps badgers are commoner in arid regions. AGE AND RELATIONSHIPS OF THE FAUNA A Pleistocene age seems clearly indicated by the assemblage, but a more critical recognition of the stage represented in the absence of stratigraphic evidence is not entirely satisfactory. Our knowledge of the sequence of mammalian forms during Pleistocene time is very in- complete, and it is only through direct correlation of fossil occurrences with known glacial and interglacial deposits that the succession of faunas can be determined. The appearance of new migrants and the extinction of old types appear as valuable criteria in recognizing various stages, but in any case the absence of forms can never be regarded as more than negative evidence. The number of extinct forms recognized in the fauna and the diffi- culties encountered in identifying as modern several of the species so classified do not favor a very late stage at Cumberland Cave. Many typical Pleistocene forms, however, as saber-toothed cais, sloth, camel, elephant, and extinct species of bison and musk-ox, are not included in the assemblage. Here again the absence of types is not positive evidence. It is a noteworthy fact that remains of animals larger than a black bear are exceedingly scarce in the collection, probably owing to the mode of accumulation and possibly m part to regional environment. Gidley (1913b, p. 95) regarded the assemblage as pre-Wisconsin in age and possibly as old as that from Port Kennedy (Cope, 1899) m Pennsylvania. A comparison of the Cumberland Cave faunal list with that of the Port Kennedy fauna, and with the list of mammals from the Conard Fissure in Arkansas (Brown, 1908), follows. 8 BULLETIN 171, UNITED STATES NATIONAL MUSEUM CUMBERLAND OAVE Sorez sp. Blarina brevicauda Eptesicus cf. grandis Corynorhinus alleganiensis Canis armbrusteri Canis cf. priscolatrans Euarctos vitabilis Arctodus haplodon Martes parapennantt Mustela cf. vison Gulo gidleyi Intra parvicuspis Brachyprotoma pristina Spilogale marylandensis Taridea marylandica Felis cf. inerpectata Felis near atror Marmota monar Citellus ef. tridecemlineatus Tamias cf. striatus Sciurus tenuidens Glaucomys sp. Plesiothomomys potomacensis Castor canadensis Peromyscus cf. leucopus Neotoma magister Parahodomys spelaeus Synaptomys cf. cooperi Synaptomys (Mictomys) sp. Microtus (or Pitymys?) cf. in- volutus Ondatra cf. annectens Zapus sp. Napaeozapus cf. insignis Erethizon cf. dorsatum Lepus cf. americanus Ochotona sp. Mammut cf. americanum Port KENNEDY Scalopus sp. Blarina simplicidens Myotis(?) sp.! Canis dirus(?) Canis priscolatrans Vulpes latidentatus Urocyon cinereoargenteus Euarctos americanus Arctodus haplodon Martes diluviana Gulo gidleyi Tutra rhoadsit EBrachyprotoma obtusata (?) Mephitis nigra Osmotherium spelaewm Pelycictis lobulatus Tazidca tarus Smilodontopsis gracilis Smilodontopsis mercerii Felis inerpectata Felis eyra Lynx calcaratus Sciurus calycinus Castor canadensis Peromyscus leucopus(?)! Anaptogonia hiatidens Sycium cloacinum Microtus diluvianus Microtus speothen Microtus (or Pitymys?) dideltus Microtus (or Pitymys?) involutus Zapus hudsonius(?) Erethizon dorsatum(?) Sylvilagus floridanus Ochotona(?) palatina Mammut americanum ' According to O. P. Hay (1923, p. 312). CONARD FIssuRE Scalopus aquaticus Sorex personatus Sorex personatus fossidens Sorex obscurus Sorex fumeus Microsorer minutus Blarina brevicauda ozarkensis Eptesicus grandis Myotis subulatus(?) Canis occidentalis(?) Vulpes fulva(?) Urocyon(?) sp. Procyon lotor Euarctos americanus Martes pennanti Mustela vison Mustela cicognanti angustidens Mustela gracilis Brachyprotoma pristina Mephitis mesomelas newtonensis Spilogale interrupta(?) Smilodontopsis troglodytes Smilodontopsis conardt Felis couguar Felis longicrus Lynx compressus Lynz rufus(?) Marmota monar Citellus tridecemlineatus(?) Tamias nasutus Sciurus hudsonicus Geomys parvidens Castor canadensis Reithrodontomys simplicidens Peromyscus sp. Neotoma czarkensis Microtus ochrogaster Ondatra annectens Erethizon dorsatum Lepus americanus(?) Lepus giganteus Sylvilagus floridanus(?) PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 9 CUMBERLAND CAvE—Con. Port KENNEDY—Con. CONARD FissurE—Con. Equus sp. Equus complicatus ! Equus scotti(?) Equus pectinatus | Tapirus sp. Tapirus haysii Platygonus cumberlandensis Platygonus(?) tetragonus Platygonus vetus(?) Mylohyus exortivus Mylohyus nasutus Mylohyus browni Mylohyus cf. pennsylvanicus Mylohyus pennsylvanicus Mylohyus sp. indet. Mylohyus sp. indet. Teleopternus orientalis Cervus sp. Cervus canadensis(?) Odocoileus laevicornis Odocoileus hemionus(?) Odocoileus cf. virginianus Odocoileus virginianus Odocoiletis virginianus(?) Bison sp.! Buceratherium (?) americanum Symbos australis Megalonyx lorodon Megalonyr wheatleyi Megalonyx tortulus Megalonyz scalper Mylodon harlani(?) The Cumberland Cave fauna exceeds that of other caves in number of mammalian genera recorded but includes a slightly smaller number of species than recognized in either the Port Kennedy or Conard Fissure occurrence. Of the 41 genera listed from Cumberland Cave 8 are extinct, and of 46 species about 28 are believed to represent extinct forms. Thirteen of the 39 genera and about 37 of the 51 species in the Port Kennedy fauna as listed are extinct. This suggests a somewhat greater antiquity for the Pennsylvania assemblage. Thirty-eicht genera are recorded from the Conard Fissure, and of this number 4 are extinct, and about 19 of the 50 species are extinct. The Conard Fissure fauna may be appreciably younger than either of the other two, although O. P. Hay (1923, p. 14) believed that the Arkansas occurrence antedated Cumberland Cave. Saber-toothed cats were found in both Port Kennedy and Conard Fissure deposits and not in Cumberland Cave, but tapir, mastodon, and Arctodus included in the Cumberland Cave and Port Kennedy occurrences are absent from Conard Fissure. It seems odd that sloth remains were not found in either Conard Fissure or Cumberland Cave. Megalonichid material is frequently encountered in caves, and species of this group of sloths probably lived until near the close of Pleistocene time. If the Port Kennedy occurrence is considered as early Pleistocene, that of Cumberland Cave may well be near middle Pleistocene in age. However, the particular stage or stages represented remains undetermined. 1 According to O. P. Hay (1923, p. 312). 10 BULLETIN 171, UNITED STATES NATIONAL MUSEUM SYSTEMATIC DESCRIPTION OF VERTEBRATE FAUNA Order LORICATA Crocodylid A single reptile tooth in the Cumberland Cave collection was recog- nized by C. W. Gilmore as belonging to a crocodile or an alligator. The importance of this tooth is seen in its occurrence farther north than the present limits of this group, suggesting a more equable climate than at present. Order SERPENTES Ophidian remains Several small vertebrae in the collection were identified by Gilmore as snake. There is no certain evidence in the material that more than one form is represented. Order GALLIFORMES Family TETRAONIDAE BONASA UMBELLUS (Linnaeus) A bird bone (U.S.N.M. no. 11690) in the collection was identified by Dr. Wetmore (1927) as the distal portion of a left humerus belong- ing to a ruffed grouse. Order INSECTIVORA Family SORICIDAE SOREX species A single maxillary fragment (U.S.N.M. no. 12468) including three teeth, apparently P*, M', and M’, was kindly identified by Dr. H.H. T. Jackson as belonging to the long-tailed shrew, Sorex. The specimen is too fragmentary to permit specific recognition, but no differences could be cited to distinguish it from the living species Sores: cinereus, which inhabits the region of the cave today. BLARINA BREVICAUDA (Say) Representing the short-tailed shrew are one rostral portion of a skull with the greater part of the dentition, two maxillary fragments with teeth, and eight lower jaws. The fossil material indicates a size somewhat greater than the average in Recent specimens and in several characters resembles the form described by Brown from the Conard Fissure as Blarina b. ozarkensis (Brown, 1908, pp. 170-171, pl. 15). The heel on the last lower molar is reduced as in the Arkansas form but not more so than in some modern individuals of B. brevicauda. The presence of this heel clearly separates the Cumberland form U. S. NATIONAL MUSEUM BULLETIN 171 PLATE 6 hotographed from near opening, showing bones sticking in clay and stalactitic material. cave interior p Small portion of ¢ PEATE BULLETIN 171 U. S. NATIONAL MUSEUM *SINIDO JIS ov punoisyorg LYS ul past dxd e}eINS dap dAvd dy] d1dYM AdTTVA dU JO SpPIs ANOS dUI UO dso sv oUIKS dt 7) I | | | [+ fo 9} | | I | out SIU] FUL “Ade A Yoo1g SI[LMA SSO198 YIIOU BUIYOO] pazVd0C] St aAvd YM ul ands 1 1] MOL A PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE Si from B. simplicidens Cope (1899, pp. 219-220). The longitudinal profile of the lower border of the mandible varies in the series of Jaws from a smooth curve, as in the Conard Fissure specimen figured by Brown, to a margin distinctly angulate below M3. The variation, however, does not appear to be greater than in living Llarina brevi- cauda, although modern specimens show a greater tendency toward a sharper angulation below M3. Furthermore, there is an average difference in the form of the large lower incisor, which in the fossil shrew is less turned upward at the point than is usual in living species. As noticed by Brown, the crowns of the teeth show pigment as in more recent specimens. Order CHIROPTERA Family VESPERTILIONIDAE EPTESICUS cf. GRANDIS (Brewn)? The large vespertilionid E’ptesicus is represented in the collection by 13 skull portions and 55 mandibular rami. About half the lower jaw specimens are exceptionally well preserved, and one pair of rami was associated with a nearly complete skull. The two best-preserved Cumberland Cave skulls (U.S.N.M. nos. 12432 and 12433) are distinctly larger than the living bat Eplesicus fuscus, in which respect they resemble E. f. grandis from the Conard Fissure. Brown considered his form a subspecies of #. fuscus, but the characters outlined by him suggest that the form might well deserve full specific rank. Furthermore, if the Cumberland Cave form is conspecific with the Arkansas Pleistocene type, as seems evident, additional characters are to be seen in the crania of the better-preserved skulls from Maryland. As in E. grandis, the Cumberland Cave form shows the following differences from £. fuscus: Skull larger, ridges and depressions on the dorsal surface of rostrum usually more pronounced, width of zygoma greater, teeth somewhat larger than average, depth of masseteric fossa slightly greater, condyle on mandible relatively larger, antero- internal basal cusp on P, better developed, and mandible of greater relative depth. Additional characters not determinable in the Conard Fissure material but present in the Cumberland specimens include a wider brain case, greater separation of the occipital condyles, and slightly larger bullae. The only difference observed between the specimens from the two localities is that the upper canine in the Maryland form is slightly smaller. However, the number of speci- mens in which this tooth is preserved is very limited in both collec- tions, and the canine in recent individuals shows a noticeable variation in size. 21 Brown, 1908, pp. 174-175, pl. 15. 16541—38——-2 12 BULLETIN 171, UNITED STATES NATIONAL MUSEUM TABLE 1.—Measurements (in millimeters) of skulls and mandibles of Eptesicus grandis and E. fuscus Eptesicus Eptesicus Eptesicus ef. grandis grandis fuscus ' A.M.N.H. U.S.N.M. U.S.N.M. no. 11795 no. 255580 no. 12432 (type) (Div. Mamm.) Distance from anterior surface of canine to occipital condyle - 18.:4)./eeon 2s eee 18.2 Width of postorbital constriction---.-..-------------------- 4.3 4.3 4.6 Width) of: brain’ casez= =5 = teen ee eee ee eee 94a aes eee 8.5 Distance across occipital condyles....--....---------------- CAN hata re eee eee 5.8 Distance across zygomatic arches. ....----.--....-=-------- TSN lea e ee see oe 12.9 U.S.N.M. no. 12438 Length of upper dentition, P4 to M3__..............------- 6.1 26.1 5.9 U.S.N.M. no. 12447 Length of lower jaw from li to condyle--.-.-.-------------- 15,2 15; 2 15.0 Length of lower dentition from C to M3-__-.-.-.------------ 8.3 8.3 8.2 1 Specimen of greater than average size. 2 Approximate. CORYNORHINUS ALLEGANIENSIS Gidley and Gazin! Figure 1 Type —Skull, U.S.N.M. no. 12412, with incomplete dentition and lacking bullae and zygomatic arches. Specific characters —Skull close in size to that of C. rafinesquia. Frontal region not so highly inflated above dorsal plane of rostrum. Median depression in the dorsal surface of muzzle not so deep. Depth of posterior portion of skull from occipital condyles to top of inter- parietal less. Occipital condyles slightly farther forward. Temporal ridges do not unite posteriorly in a median ridge. Material—Representing Corynorhinus alleganiensis are 5 skull portions and 29 lower jaw fragments. The most complete specimen is the type, which is well preserved and clearly shows the proportions of the rostrum and cranium (fig. la and 6). However, the dentition is not entire and the specimen lacks the zygomatic arches and tym- panic bullae. U.S.N.M. no 12413 is an incomplete rostral portion including all the cheek teeth posterior to the canine in the right maxillary (fig. 1c) and three cheek teeth in the left. A fragmentary third skull, U.S.N.M. no 12415, consists of the left portion of the rostrum with the teeth, P* to M’, and a part of the frontal region. The two remaining skull portions consist only of maxillary fragments and teeth. The majority of mandibles are remarkably well pre- served and include a greater part of the dentition (see fig. 1d), although only two specimens retain the incisors. Description —The skull of Corynorhinus alleganiensis is about the size of that in the living species C. macrotis and C. rafinesquir. How- ever, several characters are observed that clearly separate it from the 8 Gidley and Gazin, 1933, pp. 345-347, fig. 1. PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 13 nearest living forms. The most important of these are concerned with the less inflated condition of the brain case and appear not to be due to crushing. The frontal region in C. alleganiensis does not pro- ject so markedly above the dorsal plane of the rostrum; consequently the dorsal profile from the nasal opening to the top of the frontals more nearly approaches a straight line. Also, the median pit or depression in the dorsal surface of the rostrum is not so deep as in CO. rafinesquu and is not separated from the anterior nasal notch by so conspicuous a ridge. The height of the posterior portion of the cranium as measured from the occipital condyles to the top of the interparietal is noticeably less in the fossil. Furthermore, the con- dyles are slightly farther forward in position as indicated by the dis- tance between the preserved condyle on the left side and the glenoid fossa. This suggests that the bullae, though not preserved in the fossil, may have been of smaller anteroposterior diameter or possibly not so well separated from the condyles. The temporal ridges at the anterodorsal margins of the temporal fossae are distinct and converge backward but instead of uniting to form a sagittal crest continue posteriorly a short distance as nearly parallel ridges. This character was also clearly observed in a second, more incomplete skull portion of CO. alleganiensis (no. 12415). In all specimens of C. macrotis and C. rafinesquii examined these lateral ridges unite posteriorly to form a short median ridge. Furthermore, these ridges in C. alleganiensis are somewhat better defined anteriorly than in modern forms and at the termination of each is the very slightest suggestion of a postorbital process. Corynorhinus phyllotis Allen (1916, pp. 352-353) from the region of San Luis Potosi in Mexico, though defined as having a flattened brain case, is apparently a larger and broader-skulled type than C. alleganiensis. The rostrum of C. phyllotis is described as more depressed and having a more marked median excavation than other living species of Corynorhinus. Furthermore, the audital bullae are even larger than in the more northern species. No important or consistent differences were observed between the dentition of C. allegamiensis and that of the living forms of Coryno- rhinus. Also, the lower jaws are apparently not distinguishable from those of modern forms. Although the dental formula for the lower jaw of Corynorhinus is the same as in Myotis, not one of the many specimens from Cumberland Cave was found to represent the latter genus. It is probable that Gidley had reference to this Corynorhinus lower jaw material when he included Myotis in his faunal list of 1913. The mandible of Corynorhinus has a relatively shorter premolar series and longer molar series than does that of Myotis. The presence of Corynorhinus in the fauna is of particular interest inasmuch as it is the first record of the genus from the Pleistocene and 14 BULLETIN 171, UNITED STATES NATIONAL MUSEUM it adds a new and interesting species to this group of bats. Further- more, its occurrence in western Maryland is apparently somewhat outside the recorded distribution (Allen, 1916, pp. 338, 349-350) of the living species. The range of C. macrotis extends northward into North Carolina, and the species has been recorded from the Dismal Swamp of Virginia. C. rafinesquii is known in southwestern Virginia, Kentucky, and southern Indiana, as well as farther west. FIGURE 1.—Corynorhinus alleganiensis Gidley and Gazin: a, Skuil, type specimen (U.S.N.M. no. 12412), dorsal view; 6, same, lateral view; c, upper dentition, paratype (U.S.N.M. no. 12413), occlusal view; d, left ramus of mandible, paratype (U.S.N.M. no. 12414), lateral view. X< 3. Cumberland Cave Pleisto- cene, Maryland. TABLE 2.— Measurements (in millimeters) of skulls and mandibles of Corynorhinus alleganiensis and C. rafinesquii pallescens | to Be Corunorhinus Measurement ununor hints rafinesquii g a pallescens U.S.N.M. no. U.S.N.M. no.| 213367 (Biol. 12412 (type) Surv.) Length of skull from anterior margin of canine alveolus to supraoccipital____- 15.75 15.9 Widthtofpostorbitsl! constrictions 9-2. ees ee ee 3.8 3. 75 Depth of posterior portion of skull from inferior margin of occipital condyles to top’ of interparietal...-. =.=)... bet Tee TERE eee 5. 25 6.3 Distance from postglenoid process to posterior margin of occipital condyle _ _ 5. 65 6. 4 U.S.N.M. no. 12418 | Length of upper premolar-molar series___.....--__._.___.---_---_-________- 4.4 | 4.5 U.S.N.M. no. 12414 Length of lower dentition from C to Ma, inclusive.___..______........____ 5.9 | 5.8 | PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE = 15 Order CARNIVORA Family CANIDAE CANIS ARMBRUSTERI Gidley ¢ Figures 2-10 The type of this species is a portion of a left lower jaw, U.S.N.M. no. 7662, containing teeth, P, to My, and the alveolus for M, (fig. 6a and 6). Two paratypes were designated: A portion of a right lower jaw with teeth, P, to M,, U.S.N.M. no. 7661 (fig. 7a and 6), and por- tions of the right and left lower jaws of another individual, U.S.N.M. no. 7482, containing teeth, which include carnassials of both sides and M, and the posterior half of P, of the left side. The material in the National Museum representing this species consists of a number of well-preserved specimens (figs. 2-10), includ- ing nine skull portions of which five are more than half complete. There are also 17 lower jaws and jaw portions, one nearly complete fore limb and foot, and several unassociated limb bones and vertebrae. Most of this material was collected after Gidley published his deserip- tion of this form. The chief characters given by Gidley to distinguish Canis arm- brusteri are as follows: Size slightly less than that of C. occidentalis, as that species has been defined by Miller [1912, p. 2], but tooth characters indicate an animal quite distinct from any of the true wolves. Its principal differences are seen in the greater relative depth of jaw, smaller canine, more simple p2 and ps3, the presence of a posterior basal tubercle on p;, and in the relatively larger heel of the carnassial. The paraconid also is less expanded at base, with more perpendicular anterior face. The meta- conid is larger and higher placed, while the protoconid is less broad and full, as seen from the inner side. The carnassials as a whole suggest those of a jackal, fox, or coyote rather than those ofa wolf. * * * In the newer material is a portion of a lower jaw associated with a broken and crushed skull, U.S.N.M. no. 11881, in which most of the cheek teeth are preserved. The very close resemblance of the lower teeth in this specimen (fig. 8) with the corresponding ones of the type make its reference to C. armbrusteri logical. However, the depth of the ramus in no. 11881 is not so great. By comparison with no. 11881, other skull portions in the collection are referred to this species. Considerable individual variation exists in the series of canid spec- imens, sufficient to suggest the recognition of more than one distinct species were only the extremes represented. However, the degree of variation, particularly in size and proportions of skull and teeth, is not greater than that which was observed in a series of 35 Recent canid skulls in the collections of the Biological Survey from Gila Na- tional Forest in New Mexico. 4 Gidley, 1913b, pp. 98-102. 16 BULLETIN 171, UNITED STATES NATIONAL MUSEUM The skull of Canis armbrusteri as indicated by specimens no. 11886 (fig. 2a and 5), no. 11883 (fig. 3a, b, and c), and no. 11885 (fig. 4a, 6, and c) approximates in size that of Canis occidentalis. The rostrum is slender, but the frontals are relatively broad and heavy. The sagittal crest, though variable, is more prominent in no. 11886 and no. 11883 than in old individuals of C. occidentalis. In no. 11886 the inion projects backward to a marked degree, but not nearly so much as in the Rancho La Brea wolves. The superior first three premolars are comparatively simple and slender. The anterointernal portion of the posterior root of P? and P’ does not project lingually so much as in C. occidentalis. The upper carnassial and molar teeth are relatively large and massive. TaBLe 3.—Measurements (in millimeters) of skulls and superior dentition of Canis armbrusteri U5). | UsS: US. USS. U.S. | U.S. N.M. | N.M. | N.M. | N.M. | N.M. | N.M. Meusurement no. no. no. no. no. no. 11886 11881 11883 11885 11887 7994 Skull Length from anterior end of premaxillary to condyles_ | V270 5 Bes eee Ne aS PAO 2 a kee oa See ee Length from anterior end of premaxillary to posterior Nasalionening = 4.22.2 8 lees Se ee 140; ae Eee 1130 a oe es ee ee Length from glenoid fossae to condyles.__....--.--.--- 55) cbaes aces eases 53. 5 47 45.5 Distance from line between postorbital processes to NOT Sh Se rae ee ee ee a ee eat ae L3G [ese csoee | tener 1120 1105 100 Width of muzzle in front of infraorbital foramina...--| 151 j_._-__-- 147 eg Ate ll | a a A ls 44.2 Least width between superior borders of orbits_..._--| 156 |__------ 52.6 44.7 44.1 42 Width across postorbital processes of frontals_-__------ 167.7 | 187 167 61.5 53.9 59.6 Greatest width across zygomatic arches__......-..--.- DAGO. Fes 8225 1150 L130) 0 eee eee Width of palate between alveoli of M!__.------.-__-_- Page pe ea ae 144.5 39 38. 6 36. 5 Supertor dentition | Length of cheek-tooth series, P! to M2, inclusive_.__._| 1101 1101.5 96 06:(5) |2enescnetoeee aces C, anteroposterior diameter at base of enamel_________|..---___|-------- W3s3i [2s oe ee | SS O, transverse diameter at base of enamel_____.-...-.-_|_-------]_------- Oust sc ae SS aeeee | Le PA Vanteroposterior Gia Meterse se 5 ele ee ee ee ee | eee 8.4 Sf 6c lec accres eee Pl greatest transverse dlameters:c: 29. 45-5512 2293 oe | ee 5.5 4: Si \weres [poe see Pi anteroposterior diaméter-.--.-.---- 2-5 - e |S P44 cc oe |S ee eee ere PS greatest: transverse diameter. s2- 2 =-2 ose 5 eee DSS) | Sear ee Peet oN eee se eee eos cceae ips; anteroposterior diameters] =: 2.) = )* ese a ee eee 17. 4)) lara ites 1a 6 jes P3-oreatest. transverse diameter... - 5-2 ses eae ee 6.9 Ce ols ccet ee eee ele ee P4, anteroposterior diameter____....__-__-_-_- pane eee St 28.8 28.3 28.1} 129.4 26.7 P4, greatest transverse diameter including protocone-- 14.9 14.5 13.6 14.5 | 113.2 12.9 P4, transverse diameter across paracone__.-._._.------ 11.5 12 11.9 12 11.9 11.5 M!, anteroposterior diameter parallel to outer wall. __- 18.1 18 18, 2 18 19.1 17.2 M1, transverse diameter perpendicular to outer wall__| 21.8 22.5 22.7 21.7 23. 4 20.9 M1, greatest transverse diameter__...._.----.-.------- 24.4 25. 4 24.9 24.5 25. 5 23.4 M2, anteroposterior diameter-___....-.___.-.----_----- 10.3 10.3 O50. eae 10.4 8.3 M?, greatest transverse diameter_.._..__.-.-.--.-_---- 15.7 15 1557 pss dees 17.5 14,2 1 Approximate. Two isolated upper molars of Canis armbrusteri were described recently by Bryan Patterson (1932). He noted the large size of these 17 teeth and the relatively large M? as compared with M!. The Field Museum M! possesses a well-developed hypocone, as do several of the National Museum specimens, but, as anticipated by Patterson, this character is variable and in some dentitions where the heel of M! is narrower the hypocone is not so prominent. However, the reduction of the hypocone in these cases does not reach the stage common in C. dirus of Rancho La Brea. PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE TaBLE 4.—Measurements (in millimeters) of mandible and inferior dentition of Canis armbrusteri woe | sh | Wa, | | NS | NS Measurement wOB? no. no. no. no. no. (type) 11881 8168 8169 11887 8172 Mandible | Depth of jaw below anterior part of P3_.....---------- eee 27 dea eaeSe 25. 5 21.4 Depth of jaw below heel of M.----..----------------- 36.3 32. 6 31 31.3 29 125 Thickness of jaw below heel of Mj-------------------- | 16.3 15.7 15.6 15 14.6 12.9 Inferior dentition Length of tooth series from anterior margin of canine alveolus to posterior margin of Ma-_-----------------|-------- 2127 S130 hee [a eee | pees 1106 C, anteroposterior diameter at base of enamel-_--------|-~------ WIGS6: | BA: - oes Soe 142 C, transverse diameter at base of enamel_-__---.-------|-------- 141.3 10) = Ses cca leoseea 13 Pi anteroposterior diameter-. 22. 2-=2==-5- 22s ee | eee 7 6 Gi eae ea etree eee ese Pi sereatest transverse diameter..--.---------—22-- eae] a nee 5.2 4587 eee aioe aioe oe Ps anteroposterior diaiielerae se-=- <2 <5 ae eee oa eee 13.8 loses sae 13.9 12.4 Po, greatest trans verse) diameter-_.-_...--.----225--.2_}_-. <= --|_- =e 6:4 | esses 5.4 5.6 Ps,-anteroposterior Ginmeter=—2--+-------- oe a— ne eee |e=—onn 15.4 15.8 16 14.6 13.4 Es, greatest. transverse: diamoeter--...-=-..-.s2-2n22-2a|P oo 6.7 6.9 TAS eauaee 5.8 Pi; anteroposterior diameter222--=._... 2-22-28 17.9 17.7 17.8} 118 18.3 16 P,, greatest transverse diameter__-----.--------------- 9.1 9.2 8.8 19.5 9.6 Ponti My, anteroposterior diameter--_------------------------ 31 31 131 1294 eae 27.9 Mi, greatest transverse width of trigonid__---.-.-.---- 12.5 13.2 12.1 122) Soke 10.6 M1), greatest transverse width of talonid---.-.-------- 12.5 PAO LOS eel ae ey te ae eee oe Mo, anteroposterior diameter--.-_---------------------- 13.6 13.3 | 113.5 LSS Sl oes 12.4 Ms, greatest transverse diameter-_--.-.-.-------------- 10. 4 9.8 19,9} 110 10. 2 8.8 Mawanteropostenion Giame tener —- == ee a a ee ee a nee ee 7.2 6:5 }=22-82 5 Mis greatest transverse diameter--------=--------s2-=5|aeaa-n |= - =n en |= —-— = 5.6 PS Sele 1 Approximate. C. (Aenocyon) dirus is distinctly more robust than C. armbrusteri. The specimens from Rancho La Brea have a much wider rostrum and the carnassial teeth are more powerfully built. The larger limb bones in the Cumberland Cave collection, presumably referable to C. arm- brusteri, are equally as long but slenderer than corresponding elements of various individuals of the Brea wolf in the National Museum col- lections. The coyotelike characters in the dentition of C. armbrusteri pointed out by Gidley do not necessarily require that this form be grouped with the coyotes in the subgenus Lyciscus or Thos but emphasize the specific separation of C. armbrusteri from the living wolves of North 18 BULLETIN 171, UNITED STATES NATIONAL MUSEUM ‘SS Ci SU Dy: 2 Sy —— SS Saag <= Seats sce SSS = SS FIGURE 2.—Canis armbrusteri Gidley, skull (U.S.N.M. no. 11886): a, Lateral view; 6, ventral view of left half of palate. Cumberland Cave Pleistocene, Maryland. Two-thirds natural size. 19 PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE ‘pUBLAIBIA ‘OUGD0ISI9[qg OALO PUL[EQUIND ‘02/5 [BINIVU SPATYI-OMJ, “[EISSBUIBO IOAOT JO SAOIA [BSNJOO PUL [BI10qRT '9 ‘oynyed Jo Jey IO JO MOTA [BIJUOA ‘q ‘[[MAS JO MTA [B1OYB'T ‘D :(ESRIT “OU “IN N'S'Q) [BIsseurEO Jomo[ pue [[Nys ‘AO[PID Mapsniguisy suDQ—'g AMAA 9 ; ; Z re SS oS . 2 Te y: > = = Sp Wp, Uy, =— > --, ay = qin,” St EO OS Gi SEN) qj << oe ME /) I ‘ (ye ss WD KE. ~ fi a ANd U, amy Wy 7 ee Or = HS Ae [i tay = oe Dy, Lies ese nF D 4 MIS ie f — Westtanier= C®) : BULLETIN 171, UNITED STATES NATIONAL MUSEUM 20 , i) ) we Ry, Me ) SRDS FIGURE 4.—Canis armbrusteri Gidley, skull (U.S.N.M. no. 11885): a, Dorsal view; 6, lateral view; ¢, ventral view. Two-fifths natural size. Cumberland Cave Pleistocene, Maryland. SS y, > Lh PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 21 FIGURE 5.—Canis armbrusteri Gidley, skull (U.S.N.M. no. 7994): a, Lateral view; 6, ventral view. One-half natural size. Cumberland Cave Pleistocene, Maryland. FIGURE 6.—Canis armbrusteri Gidley, left ramus of mandible, type specimen (U.S.N.M. no. 7662): a, Lateral view, two-thirds natural size; 6, occlusal view, natural size. Cumberland Cave Pleistocene, Maryland. 22 BULLETIN 171, UNITED STATES NATIONAL MUSEUM Si Dui : j Ly; FIGURE 7.—Canis armbrusteri Gidley, right ramus of mandible, paratype (U.S.N.M. no. 7661): a, Lateral view, two-thirds natural size; b, occlusal view, natural size. Cumberland Cave Pleistocene, Maryland. FIGURE 8.—Canis armbrusteri Gidiey: Left ramus of mandible (U.S.N.M. no. 11881), lateral view. Two-thirds natural size. Cumberland Cave Pleistocene, Maryland. FIGURE 9.—Canis armbrusteri Gidley: Right ramus of mandible (U.S.N.M. no. 11882), lateral view. One-half natural size. Cumberland Cave Pleistocene, Maryland. PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 23. America. In proportions of the skull and lower jaw the fossil form is distinctly more wolflike than C. rufus (=C. frustror), which also shows points of resemblance to the coyotes. - SS one. 7 X 2S <\ \ “ va 1 s 1 N x ee ut Ape eetn Ne Be SENG ‘ CRY ee AS \ Y Tre. \ 4 we \ \ \ 4 \ \, ~~ ‘. "S ook ee x =~ Sere eee or Sere : ae Ficuke 10.—Canis armbrusteri Gidley: Right ramus of mandible (U.S.N.M. no. 8172), lateral and occlusal views. One-half natural size. Cumberland Cave Pleistocene, Maryland. CANIS cf. PRISCOLATRANS Cope A skull fragment with part of P® and P* of the right side and the root portions of M' and M? on the left shows the presence in the fauna of a coyote near Canis latrans. The anterior portion and right side of the muzzle are incomplete, but the cranium and left zygoma are well preserved. The skull is very close in size to that of C. latrans but is slightly more robust, particularly in the region of the frontals. The small, very slender, sharp-cusped character of P? and the frag- ment of P* further substantiates reference of the form to the coyote group. Direct comparison with C. priscolatrans from the Port Ken- nedy deposit in Pennsylvania or with C. riviveronis Hay from Florida is not satisfactory, because of the different nature of the materials. The reference of the Cumberland Cave form to C. priscolatrans is based on the somewhat more robust character of the Maryland form as compared with living coyotes. Measurements of the skull fragment (U.S.N.M. no. 7660) are as follows: Distance from line between postorbital processes of frontals to inion, 88 mm; least width between anterior margins of orbits, 33.2; width across postorbital processes of frontals, 51 mm; greatest width across zygomatic arches, 110 mm (last three measurements approximate). Family URSIDAE EUARCTOS VITABILIS (Gidley) Figures 11-14; Puiatr 8 The Cumberland Cave collection includes an unusually large amount of well-preserved bear material. There are skull portions of at least 25 individuals. Four of the skulls are nearly complete (figs. 12-14), and nine more are represented by rostral portions 24 BULLETIN 171, UNITED STATES NATIONAL MUSEUM including most of the teeth on both sides. The remainder consists of maxillary fragments with teeth. About 35 individuals are indicated by lower jaws. Six specimens include both rami and most of the teeth. A large portion of the separate jaws are well preserved and have nearly all the teeth. In two cases jaws and skull were associated. A composite skeleton (see pl. 8) has been mounted from the better- preserved remains and placed on exhibition at the National Museum. This mount is composed of parts belonging to at least six individuals, but a small portion of the skeleton was necessarily restored. The scapulae, left ulna, fibulae, ribs, sternum, and some of the vertebrae were not represented in the collection. The composite skeleton has received the catalog number 10304, and is composed of the following catalog portions: Skull, no. 8188; mandible, no. 10302; three posterior dorsal and three anterior lumbar vertebrae, no. 10303; left radius, no. 8126; parts of both hind limbs and right hind foot, no. 8179; and complete left hind foot, no. 10301. Ursus (Huarctos) vitabilis was described by Gidley (1913b, pp. 96-97) from a nearly complete pair of lower jaws (fig. lla and b), U.S.N.M. no. 7665. The description is as follows: About the size of U. (Huarctos) americanus, but differs from that species in (1) comparatively larger canines; (2) wider space between the anterior cheek-teeth, combined with a less wide branching of the horizontal rami in general; (3) a relatively larger symphysis, which is more sharply constricted and more flattened latterly behind the canines; and (4) longer diastema between canines and cheek- teeth. Most of the ursid material in the collection was obtained since the above description was published. A restudy of the group with this added material at hand has brought to light new characters and has furnished further information on the constancy or variability of others. In view of these conclusions it has seemed advisable to redescribe the species on the basis of characters shown by the type, U.S.N.M. no. 7665, supplemented by a series of skull portions and lower jaws. Specific characters.—Size equaling that of the living black bear, Euarctos americanus (Pallas). Upper and lower cheek teeth relatively narrow. P* large and well developed, with prominent protocone. M? large and usually does not narrow so abruptly posterior to the metacone as does the corresponding tooth of £. americanus. Anterior lower premolars more commonly present and better developed than in EF. americanus. P, well developed. All the Euarctos material in the collection is referred to E. vitabilis, which indicates considerable variation in size for this species. Several jaws in the collection compare favorably with those of small individuals of EL. americanus or with the glacial bear, L. emmonsti. The largest specimen of F. vitabilis, a skull, U.S.N.M. no. 12247, equals in size large specimens of F. luteolus from Louisiana. PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 25 The skull of ZL. vitabilis is not so robust as that of Ursus horribilis, nor is the muzzle so elongate. In the relative proportions of the skull and lower jaw the Cumberland Cave form compares favorably with living black bears of North America. The shape as well as the size of the fossil skulls varies somewhat, but no important or consistent differences can be cited to establish a division within the fossil collec- tion, or to distinguish them from Recent forms on the osseous parts alone. In the dentition EH. vitabilis shows points of resemblance to both E. americanus and E. luteolus, but in some characters it seems clearly distinguished from them. £. luteolus possesses teeth averaging some- what greater size than does H. americanus. Also, in E. luteolus the cheek teeth appear relatively wider and P* is better developed. The upper and lower teeth of E. vitabilis are relatively much narrower than in H. luteolus and average somewhat narrower than comparable teeth in EL. americanus. Upper and lower canines in the fossil form are usually large as in £. luteolus. FP! in £. vitabilts is noticeably well developed, with a prominent internal cusp placed opposite the notch between paracone and meta- cone. This tooth is much larger than in /. americanus, and in several fossil specimens it is appreciably larger than in LF. luteolus. The corresponding tooth in U. horribilis is in turn very much larger than in L. vitabilis, and in the grizzly the protocone is accompanied by several accessory cuspules. M!, other than being relatively somewhat narrower, differs little from this tooth in EZ. americanus. M7? is an elongate tooth that does not usually taper so abruptly posterior to the metacone as is common in E. americanus; however, this tooth does not reach the development and proportions seen in U. horribilis. The anterior lower premolars are more commonly present in E. vitabilis than in EL. americanus and E. luteolus and are usually better developed. P* averages larger than in the living black bears, but it is not nearly so large as in U. horribilis, nor does it show the sulcate heel, which characterizes this tooth in the grizzly. M'is a narrow tooth having a trigonid cuspate as in the ursids not trenchant as in the tremarctine forms. The succeeding molars are also narrow but otherwise resemble corresponding teeth in Recent black bears. A large amount of limb and foot material of Huarctos vitabilis is included in the collection. The various elements for the most part were unassociated. A noticeable variation in size is seen and no consistent differences were observed which would serve to distinguish the fossil from the living bear. The skull of EZ. vitabilis differs from that of Ursus procerus Miller (1899) from the Pleistocene of Ohio in having a less concave dorsal 26 BULLETIN 171, UNITED STATES NATIONAL MUSEUM surface of the rostrum and most noticeably in the smaller size of the teeth. E. vitabilis is also distinct from U. amplidens Leidy (1856) in having lower teeth of much smaller size. Upper teeth from the Conard Fissure referred by Brown (1908, pp. 183-184) to Ursus americanus may represent Huarctos vitabilis. The characters of the upper dentition cited by Brown are also seen in the Cumberland Cave bear, as the development of P*, the narrow- ness of M!, and the character of the heel of M?. In the past most finds of Pleistocene black bear have been referred to the living species Euarctos or Ursus americanus, and it seems probable that with a very limited amount of material it would be difficult to recognize as dis- tinct a species approaching the living form so closely as does EF. vitabilis. Tasie 5.— Measurements (in millimeters) of skull and superior dentition of three species of Kuarctos Euarctos | Fuarctos Fuarcios vitabili 3 5 americanus| luteolus Measurement s yee | i M.no. | M. no. | Meno. | U-S:N.M.| U.8.1.M. 8181 12247 8143 | 20. 35388 | no. 132546 Skull | Length from anterior end of premaxillaries to posterior | endiof condyles: 2-25.24 2-23 == ee ee ee OE ene esa ose eee 284 304 Distance from line between posterior surfaces of glenoid fossae to posterior end of condyles-_--.-.----------------- Ole Fae 5243S. see ee 96 101 Distance between anterior margin of orbit and anterior endsofipremaxillary 3222000 ener - ce sen eee eee 105 1 120 112 103 108 Distance from line between postorbital processes of fron- Cait; MON aes a ae enn en eee nce Diy ($n Pepsi S| Fee See 168 184 Greatest width across muzzle from outer walls of canine alveolis =< eee oes: Re SAS ee ee eae 162 74.4 | 164 62 64.8 Least width between anterior margins of orbits__-------- 74 18350 jleoc-ses= 75 71 Greatest width across postorbital processes- ------------- 102 108) 79= 2258273 105 101.5 Greatest width across zygomatic arches__--.----.-------- TZN a Bae ny) Ne ee 180 186 Width of palate between alveoli of M?_....-.------------ 47 145 45.5 50 47.2 Superior dentition Length of tooth row from anterior margin of C alveolus to | posterior surface of Ma. 2 ~~ sss ne en 1103 1118 1110 103 107 Length of tooth row P4 to M2, inclusive----------------- LOS 64 62 54. 2 59. 5 C, anteroposterior diameter at alveolus---_-.-..--------- 120 125 121.5 20.6 20. 2 C, greatest transverse diameter at alveolus_---_---------- 11.6 15 13 12 1257 Pt anteroposterior diameter: ---- + ------ == eames 13. 2 13.8 13.7 11.5 12.1 P4, greatest transverse diameter _.-.-------------------- 9. 4 10 11.3 8.2 9.2 Mi; anteroposterior diameter: -~-.-- #2 =e == eee 19.3 19.3 21 17.2 19 M!, greatest transverse diameter_....._------.----------. 13.2{ 14.4 14.6 13. 2 15 M?, anteroposterior diameter_._......--...-------------- OT Tel S3ie ty] R273 26.6 29 M3, greatest transverse diameter_....-.--.----------.---- 15.2 | 17a 15.8 15.7 17 } Approximate. PLATE 8 BUEEEDRIN 171 U. S. NATIONAL MUSEUM *WNosn Ay [euc AYE N S$ mia): ul PIQIY XS] “QAV) purfioquing WoO] (Aa]| pis) SUUGDIIA SOJILONTT ) VIG JO (a}1sod wos) UO}JI[9YS poJUNo] 7 PE Adie 9 BULLETIN NATIONAL MUSEUM iS “UN asSN]AT |v ONE NT aS A) UL porqryxay “OAR ) pue[toquiny wold f (eH 2hajpi3 ojn4)) JULIDAJOM JO UOJaJays paqunoyy Ce et [oz[eat WN PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 27 TABLE 6.—Measurements (in millimeters) of mandible and inferior dentition of three species of Kuarctos peas Euarctos | Euarctos Euarctos vitabilis americanus| luteolus Measurement U.S.N. M. no. | V-S:N-|U-8.N.| v.s.N.M.| U.S.N.M. 7665 814 % 8143 “| no. 35388 | no. 132546 (type) Mandible Length of mandible from anterior end to posterior surface Of cond yiletee ee oso = Re oe ose oe eee senacsseess|hohseeee 1197 1195 1193 210 Depth ofijdwibelow Mir: 2252222 == 222 22 aoe oe aos 41.3 41 38 36.6 40 Thickness: of taw, belowsli 22222. 22 eee we ee oe eee 16 14.6 15.5 1733 Least width across both rami at constriction posterior to Canines 230 35_ 0. 2 epee SA eee seo tedeenesese 36 41 | oceeeee 35 37.4 Inferior dentition Distance from anterior margin of canine alveolus to pos- terior'surface, Of) Mig stores boosh Ree Soto os nas 117 121 115 118.5 123. 4 Length of tooth row, P4 to Ms, inclusive___-------------- 162 67 66.5 61 69. 2 Diastema between canine alveolus and P4__------------- 136.8 33.3 31 36. 4 32.8 C, anteroposterior diameter at alveolus..__-.-_---_------- VS. 1), 20 119 120 1 20. 5 C, transverse diameter at alveolus_---------------------- 112 12.8 11 11.6 11.9 PP, ‘anteroposterionidiameters-=o-2- eee oe oo 11 10.9 9.9 9.5 9.8 Pi; greatest transverseidianietens. .-s2ss--s--——-2------=-|-22=---- 6.7 5.8 5.0 5.7 Mi, anteroposteriondiameterses---=.25)=-sa5o- ean o 17.8 20. 5 19.6 18.6 20.5 Mi, transverse diameter of trigonid-...-----.------------ 6.9 8.3 7.3 7.9 8.6 Mi, transverse diameter of talonid_...-_----------------- 8.3 9. 4 8. 4 9.3 10.7 Ms; anteroposterior diameter. —---=--~-..-..-.2-.---=.<-- 17.7 20.8 20. 5 19.5 22.5 Ms, greatest transverse diameter-_......------------------ 11.4 12.1 WS 11.8 14.5 IMs) anteroposteriomdiametersss--)--o--4-2 = se ee 1551 15.1 17.7 13.8 17.6 Mz, greatest transverse diameter_.._...------------------ 11.9 12 12.4 11.5 13.8 1 Approximate. ARCTODUS HAPLODON (Cope) A crushed right lower jaw, with the third incisor, canine, and second molar, and an isolated second molar from the left side of the same mandible, U.S.N.M. no. 8005, show the presence of a rather large tremarctine bear in the cave fauna. Several limb bones and parts of fore and hind feet, greatly exceeding in size those of a black bear and apparently belonging to one individual, are also believed to be those of Cope’s eastern cave bear. The lower jaw from Cumberland Cave has been badly distorted by crushing but appears to be slightly deeper than the jaw figured by Cope (1899, pl. 19, fig. 2) from the Port Kennedy bone deposit. Although the posterior part of the ramus, U.S.N.M. no. 8005, is not preserved, the configuration of the jaw below M, suggests that the antemasseteric pit probably extended forward nearly to a point below the posterior margin of M,. In Ursus the anterior margin of the masseteric fossa is not so far forward. 16541—38——3 BULLETIN 171, UNITED STATES NATIONAL MUSEUM 28 “‘puBlAIB ‘OU9D0}SIE[q OABH PUB[JOqUING ‘OZIS [PINVU SPITYI-OM ‘MOTA [BIOIRI “Q f0z]S [BINJVM ‘META [VSNJOIO ‘D :(999L ‘OU “JA"N’S'Q) Ueumpoeds odéy ‘ofqipueuI Jo snureI yYSII ‘(AO[PIH) syiqnz1a sopunNn—‘T] aun sere ee ere Zi} a SSS A e D Fare ee eR = TS See 4 = Poa SS — Bun = = ; NG SSX ey ete naes N SS . Seen <= : Coo ea \ v3 > o'NSSS ? GF ees LYASsg Poor Teo Kote. EES spins Z —- edhe : ae Xs ESS ssa: : ite eS ; = Mer, > =~ + ee — : typ Vint PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 29 | ZY #£ ds FIGURE 12.—Ewarctos vitabilis (Gidley): Skull (U.S.N.M. no. 8181) and mandible (U.S.N.M. no. 8141), lateral view, three-fifths natural size. Cumberland Cave Pleistocene, Maryland 30 BULLETIN 171, UNITED STATES NATIONAL MUSEUM FIGuRE 13.—Huarctos vitabilis (Gidley): Skull (U.S.N.M. no. 8181), dorsal view, three-fifths natural size. PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE My a “=S l : BD ee ), i) > ai ple DM \ Wie 4 pn ( A DE ie We Gy rs Sy YZ Sg i) ge iM if (C7E- AW f - IER? UA (es x(\ YZ : Wy LAS Ne W237 ‘As 2B iWees= Vy Ee Aiod NES ey \ Ve N/ AP ¢ A 5 G We 2 el i fa Wy = / wi! al WN LS \ 2 jy ),! inp “1 A, } | u aed vA \\ : a yj ol eS Loe I EN NG = <= Ez y Ee SS \ [“ SX 4 ‘If | ot y Y ruts MIE, ens 14° 4; re SLL = G XS wf ‘(ES a 31 SO OS SS = we “> a of IS EW Nip Bx fii \\ K> Li RNy Figure 14.—Euarctos vitabilis (Gidley): Skull (U.S.N.M. no. 8181), ventral view, three-fifths natural size. 32 BULLETIN 171, UNITED STATES NATIONAL MUSEUM The canine and second molar are nearly identical in size with these teeth in the Port Kennedy jaw. ™M, has an occlusal surface as that in Cope’s specimen, but not so tuberculate as in Arctodus pristinus Leidy ; however, the external surface is deflected as in Leidy’s specimen (Leidy, 1860, pl. 23, figs. 3,4). Arctodus floridanus Gidley (1928) is a distinctly smaller species approaching closely the living Tremarctos inornatus Gervais. The second molar from Cumberland Cave is noticeably narrower than in the western forms of Tremarctotherium. The Frankstown Cave form, referred by Peterson (1926) to Are- totherium haplodon, compares closely in size with the lower jaw from Port Kennedy. M, in our specimen is similar to that in the Carnegie Museum specimen and shows only a small difference in the outline of the external surface. In the general appearance of this tooth the Cumberland Cave form is perhaps closer to the Port Kennedy form than is the figured specimen from Frankstown Cave. The astragalus of the partial skeleton, U.S.N.M. no. 8180, referred to Arctodus haplodon corresponds very closely to the dimensions given by Cope for the Port Kennedy A. haplodon, but the third metatarsal is distinctly smaller than that measured by Cope. It seems probable that the difference is to be accounted for by size variation between individuals in the Port Kennedy fauna. The foot bones of the Cumberland Cave specimen are noticeably small compared to Trem- arctotherium simum (Cope), according to dimensions given by Merriam and Stock (1925), although the tibia has nearly the proportions of that in the Potter Creek Cave form. 1. californicum (J.C. Merriam) from Rancho La Brea is very much larger (Merriam and Stock, 1925). Family MUSTELIDAE MARTES PARAPENNANTI Gidley and Gazin § Ficure 15 Type.—Anterior portion of male skull (fig. 15a), U.S.N.M. no. 8010, carrying most of the cheek teeth of the left side. Material—In addition to the type there are 10 specimens in the collection, consisting of lower jaws and skull fragments with teeth. One pair of maxillaries, U.S.N.M. no. 12336, with nearly all the cheek teeth on both sides, has a fragmentary right ramus of the mandible associated. Specific characters.—Size intermediate between Martes pennanti and M. americana. Skull construction, as far as can be determined, nearly as in MZ. pennanti. Upper carnassial relatively shorter than in M. pennants with protocone more forward in position as in M. americana. Upper carnassial with uncovered and well-developed external median rootlet as observed in M. pennanti. Mi! relatively short anteropos- teriorly with metacone and cingulum posteroexternal to metacone 5 Gidley and Gazin, 1933, pp. 347-349, fig. 2. PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 33 somewhat reduced. External roots of M' not so distinctly separated as in M. pennanti. M, with metaconid much reduced. Mandible with low but anteroposteriorly elongate coronoid process. Masseteric fossa deeply impressed. Coronoid process with two well-defined fossae, one on its external upper half, the other on its inner base. The modifications of the coronoid region are more pronounced than in living species. TasLe 7.—Measurements (in millimeters) of skull and mandible of Martes parapennanti U.S.N.M. | U.S.N.M. | U.S.N.M. Measurement no. 8010 no. 11877 no. 11878 (type) (male) (female) Skull Width across postorbital processes-_----.---.-------.---------------- WIS) |Scceussesase |bass-oseccee . Least width between superior borders of orbits...------------------- 194) ee |S cocsecause Width between outer surfaces of M!__..---.-.---------------------- N92" ac cccebosos|sbekesossecs Superior dentition Length of cheek-tooth row, P! to M!, inclusive.-..------------------ 805i pe seeee Siac eee oo ceee P4, anteroposterior diameter--------..--...------------------------- Diu poccee cone aacomaan oe P?, greatest transverse diameter-._.--------------------------------- BG oe ecceses|sosececsenoe Pe anteroposterior GiaNleLen. = 22 ese sa neo eee eee nae Seema ena Gi Gale as Sock adee P3, greatest transverse diameter_-_...-----.------------------------- Ott | oseecne sacs | P Sse ee conan Pi anteroposterior diameter. = ---------- ae ana aan eee == BO oN wwo caetecc|-cosecceescae P‘, transverse width across anterior part including protocone-------- Gee [bccn ascent leaeeecoeso oe P4, transverse width across metacone------------------------------- Be see M1, anteroposterior diameter of outer lobe_------------------------- Gi oenccce een tees cacecaae MI, anteroposterior diameter of inner lobe_-.----------------------- 6* leeetccecee ws Eososeotoece M1, greatest transverse diameter--..-------------------------------- OUR ease sacse oes |aeaneaccosne Mandible Length of lower jaw from anterior surface of canine at cingulum to middle of posterior surface of condyle-.-.---------------------------|------------ 6825) |cceacaeaas—o Distance between posterior surface of M: and posterior margin of coronoid above condyle--------2--2---2--=--------- <<... =----=-|----=------- Bi 2 (| oe ee Height of coronoid above base of angle__----------------------------|------------ BL eenensecsase Depth of jaw below heel of M: on external side_---------------------|------------ 12.6 9.2 Inferior dentition Length of tooth series, C to M; inclusive_.--_-----------------------|------------ 40.8 136.5 Py anteroposterior Giameteres se ess sesne nee aaa eee aeee | eae ee eee enna ene a Ps, anteroposterior diameter-..-..----------------------------------|------------ 5 2.8 Ps, greatest transverse diameter--._.--------------------------------|------------ 2.6 2.1 Ps, anteroposterior diameter---------------+---=--=_-_-_=-2-<---0-|--------=__- 6.4 5.4 Ps, greatest transverse diameter-------------------------------------|------------ 2.8 2.3 P,, anteroposterior diameter_--_--.-_--------------=---..---_-------|------------ 7.9 6.4 Pi, greatest transverse diameter-__...--------------------------------|------------ 3.3 2.6 Mir aTiteropOStenlor Gia MeLeheessess sae eee ee ane enna an ee |e 12.7 110.5 Mi) transverse width of trigonide 22 oo enon eee eee eee eee ean == 4.6 3.6 My, transverse width of talonid.---_-.------------------------------|------------ Ay 7; || ee 1 Approximate. In Martes americana and M. caurina the upper carnassial is rela- tively short, as it is in M. parapennanti, but P* in these living forms 34 BULLETIN 171, UNITED STATES NATIONAL MUSEUM lacks the external median rootlet characteristic of the fishers. M1! also is relatively short anteroposteriorly in the martens and the meta- cone is reduced, more so than in M. parapennanti. The condyle of the lower jaw in M. parapennantt is relatively narrow transversely and shallower than in the living species examined. A glenoid portion preserved with one of the specimens, U.S.N.M. no. 11959, shows a correspondingly narrow glenoid fossa and the post- glenoid process curves well forward along its inner border to form an interlocking joint as in Gulo. In the living species of martens this joint is perfectly hinged but is not completely interlocking. FIGURE 15.—Martes parapennanti Gidley and Gazin: a, Anterior portion of male skull, type specimen (U.S.N.M. no. 8010), lateral view with occlusal view of dentition; b, right ramus of female mandible, paratype (U.S.N.M. no. 11878), lateral and occlusal views; c, right ramus of female mandible (U.S.N.M. no. 11876), lateral and occlusal views; d, left ramus of male mandible, paratype (U.S.N.M. no. 11877), lateral and occlusal views. Natural size. Cumberland Cave Pleistocene, Maryland. Martes diluviana (Cope, 1899) from the Port Kennedy deposit in Pennsylvania has a much shallower jaw, as indicated in the two figured specimens, than male specimens of M. parapennanti with teeth of com- parable size. The depth of the jaw is near that of a female M/. para- pennanti with teeth considerably larger. Moreover, the metaconid of the lower carnassial is apparently better developed in M. diluviana. This is noticeable when comparisons are made with U.S.N.M. no. 12352 from Cumberland Cave, which has relatively unworn teeth. There is but a very slight development of the metaconid on Mj, and this PLEISTOCENE VERTFBRATES FROM CUMBERLAND CAVE 35 specimen though exhibiting a youthful dentition has a relatively much deeper ramus than in either of the youthful specimens figured by Cope. Dr. E. R. Hall (1936), however, regards these differences as of only age significance. The modifications of the upper facial region in the Cumberland Cave form resemble those seen in M. nigripes, but in the latter species the postorbital processes are more prominently developed and the convergence backward to the postorbital constriction is more extreme. The dental characters throughout are decidedly more like those of M. vison. In M. nigripes the protocone of the upper carnassial and the protocone shelf of M! are more reduced, and the heel of the lower carnassial is very narrow, with the talonid basin entirely lacking. MUSTELA cf. VISON Schreber Ficures 16-18 Included in the mustelid material from Cumberland Cave are six skull portions and an equal number of lower jaws belonging to mink (see figs. 16-18). The fossil material closely resembles the form M. vison mink living in Maryland at the present time. A number of differences were observed between the fossil and modern forms, but the magnitude or importance of these differences is not greater than that of the distinctions seen to exist between several of the subspecies currently recognized. TaBLE 8.—Measurements (in millimeters) of skull, mandible, and dentition of Mustela vison nn nee EEUU ESSE Sa Must_la v. | Mustela ov. Mustela cf. 7 re a eae vison S.N.M. .S.N.M. Measurement U.S.N.M. | no. 2242 | no, 241192 no. 8156 (Biol. (Biol. Surv.) Surv.) Skull and superior dentition Distance anterior end of nasals to line between postorbital processes_ lant) 15 16.8 WidthubetweenlOnDitSee see esea anes ee een e nena seen a nena eee 14.5 14 14.7 Width at postorbital processes--_=-----.----...---------..------.--- 16.8 16 16.3 Width across outer margins of canine alveoli-____-------------------- 13.8 1255 13.6 Length of upper tooth series, anterior margin of canine alveolus to posterior surface of carnassial-__----------------------------------- 19 17. 4 18.4 P4, anteroposterior diameter along outer wall_-_--------------------- 7.4 6.7 7.5 P4, transverse diameter including protocone-_------------------------ 4.7 3.8 4.4 P4, transverse diameter posterior to protocone----------------------- 2.6 2.2 2.4 Mandible and inferior dentition Depth of lower jaw below trigonid of M1-.--------------------------- 6.5 6 6.7 Distance from apex of coronoid to base of angle--------------------- 17.7 17 18.4 My, anteroposterior diameter: ----------------------------=------=--=-= 8 7.5 8.2 Mi, transverse diameter of trigonid over anterior root--------------- 2.5 2.4 2.5 Mi, greatest transverse diameter------------------------------------ 3 Sal 3.2 Mi, transverse diameter of talonid...------------------------------- 2.8 2.9 3 I 36 BULLETIN 171, UNITED STATES NATIONAL MUSEUM WF Mg j FIGURE 16.—Mustela cf. vison Schreber, anterior portion of skull and right ramus of male mandible (U.S.N.M. no. 8156): a, dorsal view of skull; b, ventral view of skull; c, lateral view of skull; d, lateral view of mandible; e, occlusal view of interior dentition. x 144. Cumberland Cave Pleistocene, Maryland. PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE on In size the fossil form is close to M. vison mink. A comparison with this subspecies shows the muzzle of the fossil skulls to be rela- tively broad posteriorly with a slightly elongate face. Viewed from above the lateral lines of the postorbital constricted portion of the Figure 17.—Mustela cf. vison Schreber: Anterior portion of skull and mandible (U.S.N.M. no. 11880), anterior and lateral views. Natural size. Cumberland Cave Pleistocene, Maryland. skull (fig. 166) converge posteriorly, perhaps more so than in the Alaskan form, M. vison ingens. The upper dentition is essentially like that in M. vison mink, although the upper carnassial of U.S.N.M. no. 8156 (male) (fig. 16c), is somewhat broader posteriorly than in the living mink. The lower jaw (fig. 16d and e) is less robust than in FiGure 18.—Mustela cf. vison Schreber: Right ramus of female mandible (U.S.N.M. no. 8212), lateral and occlusal views. X114. Cumberland Cave Pleistocene, Maryland. M. vison mink, but the lower teeth, with the exception of M’, are not different. The heel and middle portion of the lower carnassial are narrower in the fossil jaws, and the hypoconid appears to be more medially placed. The relative breadth of this tooth does not vary greatly within subspecies, but between such forms as M. v. vison and M. »v. ingens the proportions differ appreciably. Incidentally, the junior author does not follow the procedure adopted by Dr. E. R. Hall (1936, pp. 114-115) in referring the fossil material to the living subspecies, M. vison mink, occupying the Maryland region today. 38 BULLETIN 171, UNITED STATES NATIONAL MUSEUM GULO GIDLEYI Hall ® Figures 19, 20; Pats 9 The normal habitat of the wolverine today is the far north, and in the Eastern United States it has not been known to range farther south than northern New York. This animal is represented in the Cumberland Cave collection by several specimens including an unusually well preserved skeleton. The wolverine skeleton, U.S.N.M. no. 8175, has been mounted and is on exhibition in the hall of fossil vertebrates at the National Museum AS y Ly), FiaurE 19.—Gulo gidleyi Hall: Skull and mandible (U.S.N.M. no. 8175), lateral view. Two-thirds natural size. Cumberland Cave Pleistocene, Maryland. (see pl. 9). The skull and mandible belonging to this specimen are nearly complete and most of the teeth are intact. The ribs and the anterior portion of the vertebral column are largely restored, and the sacrum and most of the caudal vertebrae were not preserved but the lumbar and posterior dorsal vertebrae are in good condition. The right scapula is only half complete but has been restored from the one on the left side which has very little missing. The fore limbs and right fore foot are entire. Several carpals, phalanges, and claws of the left manus were not preserved but have been modeled from those on the right. Of the pelvic girdle only the tip of the left ilium and the symphysial portion of both ischia are restored. The hind limbs are complete except for the fibular and one patella. The left pes includes most of the elements and served to model the right hind foot, which had only the astragalus, cuboid, and third metatarsal preserved. In proportions of the skull a noticeable degree of variation is seen between specimens in the cave collection. The skull (figs. 19, 20) belonging to the preserved skeleton, U.S.N.M. no. 8175, is character- ized by a broad frontal region and well-expanded zygomatic arches; moreover, the posterior narial passage is noticeably constricted, and 6 Hall, 1936, pp. 83-86. PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 39 the pterygoids approach each other somewhat more than in the living species. However, skull U.S.N.M. no. 8176, the type as designated by Hall, in the cave collection, is not so broad through the muzzle, and the posterior narial aperture is not different from that in an average Z va Mi Miwa loss f MWA i XG ' ‘ LN) Ge ANA asttling tT Se N\ We i) Fiaure 20.—Guwlo gidleyi Hall: Skull (U.S.N.M. no. 8175), dorsal and ventral view. Two-thirds natural $128. modern specimen of Gulo luscus. The occipital condyles in no. 8175 are close together, and the foramen magnum is correspondingly small. Unfortunately the occipital region is missing in other skull portions from Cumberland Cave, but the distance between the condyles in living specimens of G. luscus varies sufficiently to include the condition seen in no. 8175. 40 BULLETIN 171, UNITED STATES NATIONAL MUSEUM TABLE 9.—Measurements (in mallimeters) of skeleton of Gulo gidleyi Measurement a aire : Skull Total length of skull from anterior end of premaxillaries to occipital condyles__.._----------_- 136 Widthiat:mastoids_02 "2. =!: 202) 502 eee ence ee eek ee ee ee 77.6 Widthiacross'zygomatic' arches... 2222-22-22 5 ee eee 98. 8 Width'ofimuzzleiat' canines= +=. - 2-250 =e See SS ee es ee ee 37.5 Width of muzzle atjinfraorbitalforamensss. "loos foe 2S ee ee ee 41 Width: between orbits: ==> 2S ue. Sot See eS Se ee 43.3 Width/across: postorbitaliprocesses==e2-- soo a ee ee ee ee 50. 6 Width:at postorbitalconstriction_2-— 22-2. Ao ee ee 37.2 Superior dentition Length of upper tooth series, canine to M!, inclusive. ..---....-....-.---------.---.----.----- 53.5 Ps anteroposterior diameter. - asa--2osee oon nonne nen eeate= sane an eas ee ee eee a 18.9 P4, transverse'diameter including protocone-j.=2=- 22 = 2s conc nce nn see ees cce sna eee eee ll P4, transverse diameter (greatest) posterior to protocone..-.-.-.......-.....------------_.-.-- 8 M1, anteroposterior diameter of external portione.2.- 22... 2-2 ooo 22a see ee ee 7 M1 anteroposterior diameter’ of internal) portion==22- 2202s 2ee oe seccccesaceeceeneso eae 1.21 Mi, transverse: diaméter=..- 322.2220) o es edn das eae cee atnesas anceaneeeee ee eee ee 12.5 Mandible and inferior dentition Depth of lower jaw: below trigonid of; carnassial_—-----2. _-... ees eee 19.5 Length of lower tooth series, canine to Ma, inclusive._..--.-..2.-----<_--.-.----=..<-+-------- 60. 5 Mi:;, anteroposterior diameters. 2.25282 Senet 8S soese cae so oee tow sae atccescenea ee ees 19.5 Mi, greatest transverse diameter of anterior portion. --_-.......--.-.----.-------------------- 8.4 Mi, greatest transverse diameter of posterior portion. —_-.......-.-...--...------------------- Ton Anterior limb Greatest length of scapula from coracoid process to suprascapular border___-.--.-..---------- 191 Greatest width of articulating end of scapula measured across glenoid cavity___-.--------.-_- 26.5 Distance from inner border of glenoid cavity of scapula to outer surface of acromion_____-__-- 26. 2 Greatest lengethofshumoerus #2525 226 2s eae es a eee Sone ee een 128 Greatest transverse diameter of proximal extremity of humerus__._--...__-___-------_--_______ 27.2 Greatest anteroposterior diameter of proximal extremity of humerus_......---.-.---_-------__ 20.3 Greatest width of distal extremity: ofishnmerns2...-2 == 2-2 22~ 5 aoe ne ne ee 34.6 Greatest length: of ulna 2e2: = 22. 2 See os a oa ee ee 129 Greatest transverse width across sigmoid cavity of ulna.._.-.-........-------------_-- 22 15.3 Distance from distal margin of sigmoid cavity to distal end of ulnma__.___--__----.-______.____ 104 Greatest length: ofraditis’..25 2225 iv oso ee a es 107 Long diameterrofproximaliemd: Of radi sso see es ae a a 15.8 Greatestiwidthrofdistaliend offradius sos see ease ee ee ee ee 20 Length of first moetacarpal:: sick See ates 2s Re ee ee ee 25.9 Lenethiof/second;metacarpal ses ses. see re ee oe ee ee es 35.9 Length of'third:metacarpal sss can cs aoe tee ee ee ee ee ee eee 44.5 Length:.of fourth metacarpali 2.2 "si s<200 eee ee pn ee eee ee ae 42.8 Length of fifth metacarpal: 2-2! 6.. Sarco! es ee ee ee 37 Posterior limb Length from anterior end of ilium to posterior border of ischium__-_---.-.--...-----.---------_ 112 Width ofilitnm2-0) 102 eres aren SMe 8 a os SO a a ee ee ee ee ee 30 Width of ischium measured from ischial tuberosity to posterior end of ischial symphysis_----- 45 Diameter of acetabulum measured at right angles to long axis of internal notch______________- 18.5 Length of femur from top of head to distal margin of inner condyle_-_----------------------__ 134 Width of proximal end of femur from inner surface of head to outer surface of greater trochanter_ 34.4 Anteroposterior diameter of head offers sss ee oe eee 16.4 Widthiofidistal‘end of femur‘across cond yles¥4- 2. ae ee Se 30 Greatest length of tibia measured parallel to long axis.___.--.......-..----.-.---------------- 133 Greatest anteroposterior diameter of proximal end of tibia_____..-......___._-.--__---________- 27.8 Greatest, widthiof proximal end Jofitibia 22 te Ee ee eae 29.4 Greatest transverse width of distal end of tibia measured perpendicular to long axis of shaft __ 21 Length oftfirst metatarsal’ 32.2. a). Vd 0 eee ee ee eee ee ee 34.4 Length/of'sécond metatarsal-2) 15 Soft Baird Vat hi eee bi Ot Pi ea 1 44.6 Length of third metatarsal 20 =-- 28>). Le een ee ee ee ee 52. 5 Lengthiofifourthimetatarsal vaste. ose a a eae a eg 54.2 Lengthiof fitthi metatarsals ii1f 44) visti BOP Bam See Ge eG 51.5 1 Approximate, PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE Al The teeth average somewhat smaller than in Gulo luscus, as noted by Hall, although a few specimens were found in which the teeth were definitely smaller than in the fossil form. The relatively wide P? in the fossil material was matched in some recent individuals but this condition is uncommon in CGulo luscus. The other bones of the skeleton, no. 8175, show but few peculiarities, the principal ones being that the anterosuperior border of the scapula is more rounded than in the living species, and the shaft of the humerus is slightly less curved. The record of Gulo in the Pleistocene of North America is very incomplete, and the material for the most part has been referred to the living species, Gulo luscus. Several lower jaws of wolverine in the Port Kennedy occurrence, believed by Cope (1899, pp. 229-230) to be specifically identical with Gulo luscus, are referred to Gulo gidleyi by Hall. LUTRA PARVICUSPIS Gidley and Gazin’ FiaurE 21 Type—Portion of a broken and crushed skull (fig. 21), U.S.N.M. no. 8213, in which the principal cheek teeth of both sides are present. P!, the canines and incisors represented by their alveoli. Specific characters—Size somewhat larger than Lutra canadensis. Main outer cusps of M! relatively lower and less robust than in L. canadensis. P2 and P® large with conspicuous posterointernal basin. P‘ triangulate, with its base relatively wide anteriorly and with its anterior border forming a right angle with the outer border of the tooth. P*long and outer wall of M' relatively short anteroposteriorly as compared with L. canadensis. M’ with paracone and metacone subequal and both cusps with relatively narrow extension of the basal portions of their outer walls. Comparison.—The Cumberland Cave specimen represents an extinct species of otter intermediate in size between L. canadensis of the Eastern United States and L. felina or L. paraensis of Chile and Honduras, respectively. Judged by their alveoli the canines of the Cumberland Cave specimen are relatively robust, as in the living species of Central and South America. The backward extension of the protocone shelf of the carnassial and the posterointernal shelf of the other premolars is somewhat more expanded than in L. canadensis. The posterior margin of the upper molar, as in some South American specimens, is not so deeply notched posterointernal to the metacone, as noted in L. canadensis. Perhaps the most noticeable distinction in the dentition of L. parvicuspis is the distinctly narrow proportions of the external cusp row of M’. Although L. parvicuspis is distinct from both North American and South American otters, it approaches living species of South America 7 Gidley and Gazin, 1933, pp. 349-351, fig. 3. 42 BULLETIN 171, UNITED STATES NATIONAL MUSEUM in several of those characters in which it differs from northern forms. Most noticeable among these are the size of the skull and canines, the development of the upper premolars, and the somewhat smaller relative size of the molar. FIGURE 21.—Lutra parricuspis Gidley and Gazin: Maxillary portion of skull with premolar and molar teeth, type specimen (U.S.N.M. no. 8213), lateral and occlusal views. Natural size. Cumberland Cave Pleistocene, Maryland. A comparison of L. parvicuspis with L. rhoadsii Cope (1899, pp. 238-239) from Port Kennedy, Pa., is difficult on account of the different nature of the materials preserved. L. rhoadsii was based on lower jaws and a single upper molar. The dimensions given by Cope for the anteroposterior diameter of the lingual portion of M! shows this tooth to be much smaller than in LZ. parvicuspis. TaBLEe 10.—Measurements (in millimeters) of superior dentition of three species of Lutra Lutra Lutra Lutra parricuspis canadensis felina Measurement U.S.N.M. U.S.N.M. U.S.N.M. no. 8213 no. 197943 no. 35377 (type) (Div. Mamm.)| (Div. Mamm.) Length of cheek-tooth series, P! to M!, inclusive_ _-.-.__--- 134.5 30 33.6 C, anteroposterior diameter of alveolus__._--_-------------- 17.7 6.5 7.6 ‘Pi anteroposterior, dismoeter=_ ---. 2-22 =---= =— a ee 5 5 5 P2Noreatest transverse: Glameter-.so-- se. eee 4 a3 4.3 Ps anteroposterior diameter. <225 2s 22 ee eee qed 7.5 8.2 P3: greatest; transverse diameter=.2...--.os2<5---5-8-—-=-25- 5.3 5 6 Pt anteroposterior @iameten- =. =e eee ee 12.8 11.5 14.5 P4, greatest transverse diameter perpendicular to outer wall _ 10. 2 9.6 11.5 M!, anteroposterior diameter across outer cusps__---------- 8 8.7 8.3 M!, anteroposterior diameter across inner portion____-_____ 8.7 8 8.5 M!, transverse diameter across paracone and protocone- --- 10.5 ll 11.8 1 Approximate. Satherium piscinaria (Leidy, 1873, pp. 230-231) from the late Pliocene or early Pleistocene of Idaho is a considerably larger form as indicated by the size of the tibia. (See also Gazin, 1934, pp. 143-149). E. R. Hall (1936, pp. 75-77) regards Lutra parvicuspis as a synonym of L. canadensis lataxina. In examining all the material of this recent subspecies in the National Museum the junior author finds that the PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 43 fossil species is probably distinct from the living form and that several of the characters cited above for the fossil type are distinctive and others are present only in unusual specimens or extreme conditions of the modern form. BRACHYPROTOMA PRISTINA Brown # Figures 22-24 Included in the musteline material from Cumberland Cave are three skull portions, one of which is nearly complete (fig. 22), and four mandibular rami representing an extinct form of skunk. FIGURE 22.—Brachyprotoma pristina Brown: Skull (U.S.N.M. no. 8155), dorsal and lateral views with occlusal view of dentition. 144. Cumberland Cave Pleistocene, Maryland. = The skull material corresponds closely in size and other characters with the type of Brachyprotoma pristina, which was described by Brown from the Conard Fissure in Arkansas. Skull U.S.N.M. no. 8155 (fig. 22) is rather short and broad, more so than in the living species Spilogale ambarvalis. The width is most noticeable through the frontal region and is accompanied by a reduction in size of the postorbital processes. Prominent dorsolateral swellings in the vicinity of the postorbital constriction emphasize the breadth of the 8 Brown, 1908, 176-179. 16541—-38—-4 44 BULLETIN 171, UNITED STATES NATIONAL MUSEUM skull in this region. The cranial portion is conspicuously shortened anteroposteriorly and is convex dorsally in a longitudinal profile somewhat as in the Florida S. ambarvalis. Some of the modern western skunks, as S. gracilis and S. phenaz, have relatively flat dorsal profiles and are also more elongate than the cave specimen. Further- more, the longitudinal profile of rostrum is slightly concave anterior to the frontals, as noted in several specimens of S. putorius. The dental formula of Brachyprotoma is as in Conepatus, and the characters of the upper teeth of the Cumberland Cave specimen corre- spond rather closely with those of B. pristina. M1! has a broader heel than in the type, but the difference may not be greater than can be attributed to individual variation or possibly to difference of sex. FIGURE 23.—Brachyprotoma pristina Brown: FIGURE 24.—Brachyprotoma pristina Brown: Left ramus of male mandible (U.S.N.M. no. Right ramus of female or young male mandi- 12045), lateral and occlusal views. X 1}. ble (U.S.N.M. no, 8214), lateral view. X 1%. Cumberland Cave Pleistocene, Maryland. Cumberland Cave Pleistocene, Maryland. A comparison with Spilogale shows P* in our specimen to be similar to that in the living species but with protocone relatively smaller and the anterior margin of the paracone slightly more backward slop- ing. M1! is relatively short anteroposteriorly and wide transversely, and the median constriction is not nearly so pronounced as in Spilogale. The external cingulum of this tooth is very well developed and termi- nates anteriorly in a prominent cusp external to the paracone. Of the four jaws referred to Brachyprotoma pristina, one (fig. 23), U.S.N.M. no. 12045, is evidently of an adult male and corresponds very closely to the lower jaws Brown recognized as belonging to B. pristina. Jaw no. 12046 is apparently of an adult female and corre- sponds favorably with the type of B. spelaca. The remaining two jaws (fig. 24) seem to represent relatively young males. The adult male specimen, U.S.N.M. no. 12045, is characterized by a moderately robust mandible with a rather deere abrupt symphysis, and a well- developed masseteric fossa as compared to Spilogale. The premolars are crowded, overlapping, and the lower carnassial is narrow and short heeled and exhibits a weak metaconid, differing markedly in these respects from living forms. The female specimen, no. 12046, differs from the male in having a weaker mandible, particularly Srallene be- neath the posterior portion of My, a shallower symphysis, and a more gently sloping ascending ramus. PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 45 It seems evident from a consideration of the skeletal characters distinguishing males from females among living mustelines related to this fossil form that those characters that have been cited to separate B. spelaea from B. pristina should be considered sexual. B. obtusata from Frankstown Cave in Pennsylvania, as recorded by Peterson (1926, p. 285), possibly differs from the type of that species, which is from Port Kennedy, in the size of the lower carnassial. The lower jaw illustrated by Peterson, however, does not resemble closely the specimens of B. pristina from Cumberland Cave. The difference is most noticeable in the greater downward projection of the symphy- sis, the lower position of the ventral surface of the angle, and in the differently shaped coronoid process of the Frankstown Cave jaw. If the material from Frankstown Cave should be regarded as belonging to the same species as that from Conard Fissure and Cumberland Cave, it seems preferable to refer it to B. pristina, since no comparisons are possible with the type of B. obtusata, which is lost. Cope’s meas- urements for the latter, though possibly in error, indicate a smaller species than B. pristina. Moreover, there is no assurance that the Port Kennedy deposit is the same age as that at Frankstown Cave. TasiE 11.—Measurements (in millimeters) of skull, mandible, and dentition of Brachyprotoma pristina Wi. tho) S046 | Wat | SAE | Wise | SE : .M. .H. n0. -H. .M. SHS -.M. BEES Measurement no. 12426 no. no. no. no. no. 8155 | (type) | 11722 12045 11773 12046 12399 Skull and superior dentition Length of skull from anterior end of pre- maxillaries to occipital condyles_------------ Bark YS ARSE | Seana ss | Sets ea ee eo ee Width betweenerbits_ .-.--.----------=.---== DRT S52] pe 8 ann eee oe eee ee | ee ee ee ol eee tated Outside measurement across upper molars_---} 20.1 DOS Bis eco = ee es ee SO lags ese ee ee Length of tooth series, P3 to M!, inclusive----- 1S 2tleeeee = = QR Looe eee eee ee eee eee P4, anteroposterior diameter___-----.--------- 6. 2 6.9 Gi 2)y | eee sees [Leese aCe 2 oes ree es P4, transverse diameter including protocone-- 3.9 3.8 BEO) esate | eee |oaee en |aaeeceas M1, anteroposierior diameter of outer portion of tooth parallel to outer wall_-_------------- 4 3.8 oe ee ee) pecs = el |S eee ed ee M1}, anteroposterior diameter of heel_---_------ 3.8 3 BoB u eee seers eae | ee ee eee M1, greatest transverse diameter------------- 6.3 6.7 Tee | Senn eae ee een | sen ne soem Mandible and inferior dentition Length of lower jaw from anterior end to con- iy One eae eo nee ene ee aa eat eae ae omens |[Pmmee 130 8055) |zeacees 26.6 Depth of lower jaw below P2-----------------|--------|--------|-------- 6.8 6.8 5.9 16 Depth of lower jaw below talonid of Mi------|--------|--------|-------- 5.1 5.3 4.1 4 Length of premolar series_...-----------------|--=-----|--------]-------- 6.3 6.5 633) |e See My, anteroposterior diameter-__--------------|--------]--------|-------- “se tok 7 6.8 My, greatest transverse diameter--.-.----------|--------|--------|-------- 3 3 3 2.7 My, transverse diameter of talonid_.---.------|--------|--------|-------- 2.7 2.8 2.7 2.5 ee 1 Approximate. 46 BULLETIN 171, UNITED STATES NATIONAL MUSEUM SPILOGALE MARYLANDENSIS Gidley and Gazin ® Figure 25 Type.—Right ramus of mandible (fig. 25), U.S.N.M. no. 12048, with canine and P; to M, preserved. Specific characters—Mandible about the size of that in Spilogale putorius. Symphysis abrupt and inferior margin of mandible less convex longitudinally. Canine slender and long. Third and fourth premolars and molar slender and not overlapping. P3 small. Basal portion of P* nearly oval in cross section as seen in dorsal view. Trigonid of M, relatively short and narrow. Metaconid moderately well developed but not so distinctly separated from protoconid as in FIGURE 25.—Spilogale marylandensis Gidley and QGazin: Right ramus of mandible, type specimen. (U.S.N.M. no. 12048), lateral and occlusal views. 144. Cumberland Cave Pleistocene, Maryland. S. putorius. Buccal surface of talonid not offset inwardly to so great an extent as in living species. Comparison.—Spilogale marylandensis compares favorably in size and depth of jaw with female specimens of S. putorius. However, the symphysis of the mandible is somewhat more abrupt than in either male or female specimens and projects downward to a noticeable extent giving the lower margin of the ramus a less convex and more irregular longitudinal profile. The premolars are narrower and even less crowded than is common in the living species. The trigonid of M, is slightly shorter and narrower than in either S. putorius or S. in- terrupta, and the external wall of the heel is not offset inward so abruptly from that of the trigonid as in the living species. Further- more, the metaconid of M, is more solidly united with the protoconid; however, this cusp is not reduced as in Brachyprotoma. Spilogale marylandensis does not differ greatly in size from Brachy- protoma pristina but has smaller, more slender premolars and a larger carnassial. Other characters separating the two forms are those which distinguish Brachyprotoma from Spilogale, such as the crowding or overlapping of the premolars and smaller size of heel and more reduced metaconid of the lower carnassial. 9 Gidley and Gazin, 1933, pp. 351-352, fig. 4. PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 47 Taste 12—Measurements (in millimeters) of mandible of Spilogale marylan- densis and S. putorius a . Spilogale Spilogale rinraldnateceta (panes, 0 patente Measurement U.S.N.M. no. 42892 ewe 43719. no. 12048 s E (type) (male) (Biol. (female) urv.) (Biol. Surv.) NB 8 Depth of ramus below Ps---.------------------------------- 5.5 6 5.5 Depth of ramus below posterior portion of My-------------- 4.4 5 4.4 Length of dentition, C to Mu, inclusive-------------------- 18.3 18.5 18.1 P3, anteroposterior diameter ------------------------------- 2.6 2.8 2.9 P3, greatest transverse diameter---------------------------- ere 2.1 1.9 P,, anteroposterior diameter ------------------------------- 3.4 3.6 3.3 Pu, greatest transverse diameter---------------------------- 2a 2.4 2nd Mj, anteroposterior diameter_------------------------------ | 7.4 7.4 7.6 My, anteroposterior diameter of trigonid ------------------- 4.4 4.7 4.8 Mi, greatest transverse diameter of tHigonidsese =e ne One 3.5 3.4 M1), transverse diameter of heel ---------------------------- 3 ool Sal Ne eee ee TAXIDEA MARYLANDICA Gidley and Gazin 1° Ficures 26, 27 Type.—A complete skull and lower jaws (figs. 26, 27), an articulated series of 11 vertebrae in which all the cervicals and four dorsal are present, and the proximal half of the night humerus of one individual, U.S.N.M. no. 7990. Specifie characters.—Size large, about equaling the largest living race of Taxidea taxus. Cheek teeth relatively large. Upper carnas- sial with both protocone and hypocone having low regularly cone- shaped contours with circular bases; hypocone relatively large. M1? complex in structure, with the five principal cusps (paracone, meta- cone, protocone, hypocone, and metaconule) low, regularly cone- shaped and subequal. This tooth in consequence has a flatter appear- ance than is usual in the living species. It is also more expanded posterointernally, so that the general outline of this tooth is not triangular, as is usual in the living species, but subquadrangular with a relatively straight posterolateral margin and a posteromedially directed lingual margin. Also, the posterior portion of the tooth is distinctly wide transversely. Lower P, is relatively long with well-developed heel, and the cusps of M; are distinctly robust. The skull, in general appearance, is much like that of 7. taxus, but it presents a few peculiarities that seem to distinguish it from any living species. Viewed from the side, the region above the orbit is high-arched; the occiput is relatively depressed ; and the postorbital zygomatic process is relatively low. In the basicranial portion of the skull the bullae appear to be more swollen and more evenly rounded than is usual in the living species. In superior view the principal differences to be noted are the slight development of the postorbital 10 Gidley and Gazin, 1933, pp. 352-354, figs. 5, 6. 48 BULLETIN 171, UNITED STATES NATIONAL MUSEUM processes of the frontals as compared with recent skulls in the same stage of maturity and especially the form and size of the nasals. These are broad in midregion so that the outer sides run nearly parallel for a considerable distance instead of converging \-shaped. ay ae 3 W\\) UT rs. Wy W ij ) AS S\N yi YY 1 HY FIGURE 26.—Taridea marylandica Gidley and Gazin: Skull and mandible, type specimen (U.S.N.M. no. 7990), lateral view. Three-fourths natural size. Cumberland Cave Pleistocene, Maryland. FIGURE 27,— Taridea marylandica Gidley and Gazin: Skull, type specimen (U.S.N.M. no. 7990), dorsal view... Three-fourths natural size. Cumberland Cave Pleistocene, Maryland. backward as is usual in the living species. Thus the points of the anterior extension of the frontals are very definitely farther apart than in any specimen of the living species observed. The junior anthor dees not regard this specimen as sufficiently typical of Taxidea taxus taxus to be placed in the modern species and subspecies, as does Hall (1936, pp. 79-82). Badgers today are considered strictly western in habitat, and in this country they are now not found in any numbers east of the Mississippi River. PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 49 Badgers from the Pleistocene of North America are relatively little known. Tavzidea robusta, described by Hay (1922, pp. 631-632) from Arizona, is known only from an ulna and part of an innominate bone. This material is not that in which one would expect to find significant characters of specific importance; hence there is little justification in referring it to a living subspecies, as does Hall (1936, pp. 82-83). The Anita, Ariz., material should have been cited as Tazidea sp. A specimen of 7’. tarus, originally described by Cope (1878, p. 227) as T. sulcata, is recorded from the Pleistocene of Washington. Cope (1899, p. 239) also recognized the modern species in the Port Kennedy fauna. TapLE 13.—Measurements (in millimeters) of skull and dentition of Taxidea marylandica and T. taxus Taridea marylandica fe Measurement U.S.N.M. US an Taxidea tarus no. 7990 (Div. Mamm.) (type) Lee eS SS SS Skull Total length of skull including condyles-.--------------------------------- 129.3 130 Width across mastoids.--- = <= oo eee = ne ene === = 77 73. 6 Width across zygomatic arches-------------------------------------------- 81.2 78.7 Width of muzzle across canines__.....-...----------..-----------------_-_- 36.3 35. 7 Width of muzzle at infraorbital foramina-.------.------------------------- 34. 2 34.3 Widths between O©bDitS sa seeeseseanee esa ooo eae e ne ween 30 27.5 Width across postorbital processes- ---------------------------------------- 33 33.8 Width across postorbital constriction _------------------------------------- 30 30 Depth of skull at bullae._-_--.-------------------------------------------- 51.5 56.5 Depth of occiput ---------------------------------------------------------- 42.5 45.5 Width of nasals at posterior points of premaxillaries_-_--------------------- 10.7 10.7 Width of nasals at anterior points of frontals_...--------------------------- 10.7 6.5 Superior dentition P4, anteroposterior diameter along outer wall------------------------------ 112.5 12.2 P4, transverse diameter perpendicular to outer wall_----------------------- 110.5 10.8 M1, greatest diameter-___.-------------------------------------------------- 114 14 M1, transverse diameter perpendicular to postero-external walle s2sscs 2. 110 9.4 Inferior dentition Length of lower tooth series, C to M2, inclusive._-------------------------- 60.5 52.5 Length of lower cheek tooth series P2 to M3-------------------------------- 40 41.2 Pu, anteroposterior diameter_---------------------------------------------- 9 7.9 Mi, anteroposterior diameter---------------------------------------------- 14.3 15 Mi, anteroposterior diameter of heel --------------------------------------- 5.2 5.2 1 Approximate. Family FELIDAE FELIS cf. INEXPECTATA (Cope) Figure 28 A relatively small quantity of cat material is included in the col- lection. An upper carnassial and a fragmentary left mandibular ramus (fig. 28a and 6), U.S.N.M. no. 11890, with P, and M, preserved 50 BULLETIN 171, UNITED STATES NATIONAL MUSEUM are referred questionably to Felis inexrpectata (Cope, 1899, pp. 247- 249). The proportions of the upper carnassial are such that the tooth may be from the same individual as the lower jaw." Cope described F’. inexpectata with an upper carnassial from the Port Kennedy deposit as the type. To this species he referred a lower jaw fragment with M, and part of P,, some separate upper and lower teeth, foot bones, and limb fragments. The Cumberland Cave upper carnassial is slightly larger and higher crowned than the type of F. FIGURE 28.— Felis cf. inexpectata (Cope), upper carnassial and left ramus of mandible (U.S.N.M. no. 11890)? a, Lateral and occlusal views of upper carnassial; 6, lateral and occlusal views of mandible. Two-thirds natural size. Cumberland Cave Pleistocene, Maryland. inexpectata, but it has its protocone noticeably reduced as in that species. However, the Port Kennedy lower jaw is much deeper and has larger teeth than the Cumberland Cave jaw portion and the notch between the paracone and protocone of the lower carnassial is deeper and more open than in M, of the latter specimen. The Port Kennedy lower jaw is out of proportion with the type upper tooth, which sug- gests considerable individual variation or possibly that another species is represented. The Cumberland Cave upper carnassial is nearly equal in size to the type upper carnassial of Felis longicrus Brown (1908, p. 178). The principal differences to be noted include somewhat narrower propor- tions, a more backward sloping parastyle, relatively shorter antero- posterior length of paracone, and a more reduced and less forward projecting protocone than in F. longicrus. The P, referred by Brown to F. longicrus is similar to that tooth in the Cumberland Cave man- dible but is distinctly larger. Brown noted the resemblance of this tooth to I’. onca and to the P, in the jaw referred by Cope to F. inexpectaia. The mandible from Cumberland Cave is distinctly pumalike in character, particularly in the relative proportions of P, and Mj. Though smaller than an average size jaguar, it is apparently larger 1t The specimens were associated, according to Dr. Gidley’s notes. PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 5l than the living pumas. It approximates in size F. daggetti Merriam (1918) from the Pleistocene of Rancho La Brea but possesses notice- ably slenderer teeth. TaBLE 14.—Measurements (in millimeters) of mandible and dentition of Felis cf. inexpectata Measurement aN Superior dentition P4 greatest anteroposterior diameter--—------.---2_----2-__-__--2 == -- en 26. 6 P4, anteroposterior diameter of paracone between notches------------------------------------ 10.1 P‘, anteroposterior diameter of metacone-----.....------------------------------------------- 10.9 P4, transverse diameter including protocone -_--_--.----------------------------------------- 12,2 P4, greatest transverse diameter posterior to protocone---------.----------------------------- 10 P4, height of paracone above base of cingulum externally --..----.---------------------------- 13.5 Mandible Distance between posterior margin of M: and posterior surface of condyle-------------------- 64.8 Depth of lower jaw below cingulum of M, externally-__-------------------------------------- 28 Inferior dentition PoP eaTILCEOPOSUCTIOG AIRING GIs eae se a ee 17.2 P,, transverse diameter of’anterior portion— —--------_-=_---=--_---._-_--_-_ =e 7.6 Pi greatesm@trans verse) GiamMeter a= ene e eo cere nee enn ee a ee 9.1 P4, height of principal cusp above base of cingulum externally-----.-------------------------- 12.5 IVEs sin LELOWOSLELION Ci aln Chek see an ee a ee er ee a ea 20.9 Mh, greatest transverse diameter at cingulum___.._-_._____-_.__--.-----_--_---------_------- 9.6 Miz, lenethoficutting edge of paraconid_-_-----------_- <<< ooo en _— 7.5 IMiplenet rot Cuttin ed CO lO LOL OL CCTM ee eee er eee 9.5 Mj, height of paraconid above base of cingulum anteriorly ----------------------------------- 9 M,, height of protoconid above base of cingulum posteriorly--_------------------------------- LGA FELIS near ATROX Leidy FicurE 29 Several feline foot bones in the collection appear to be much too large and robust to belong with the upper tooth and jaw portion re- ferred to Felis inexpectata and may represent the large North American Pleistocene cat Felis atrox Leidy (1853). Included in this foot material are a scapholunar and second metacarpal (fig. 29b) representing the manus and an astragalus, caleaneum (fig. 29a), and the third (fig. 29c) and fifth (fig. 29d) metatarsals from the pes. There is no evidence to show that these bones are from the same individual. These elements are much too big to belong to a large jaguar or even an unusually large puma. The proportions are comparable to those in the foot of a specimen of Felis tigris or F. leo. ‘The scapho- lunar is distinctly larger than in any of the tiger or lion carpi examined and shows a relatively deeper groove for its articulation with the magnum. However, the second metacarpal, though equally as long as in these large Asiatic and African cats, is somewhat slenderer, and the proximal end does not have so large an articular facet for the trapezoid. 52 BULLETIN 171, UNITED STATES NATIONAL MUSEUM The Cumberland Cave astragalus and calcaneum are of about the same size as in an adult Felis tugris, although the posterior portion of the calcaneum is relatively short and the proximal end is noticeably heavy. The posterior dorsal surface of the calcaneum is not so sharply 7] A y \ WS SSS a FIGURE 29.—Felis near atror Leidy, foot bones (U.S.N.M. 12840): a, Astragalus and caleaneum, anterior view; 6, second metacarpal, dorsal and proximal views; c, third metatarsal, dorsal and proximal views; d, fifth metatarsal, dorsal and proximal views. One-half natural size. Cumberland Cave Pleistocene, Maryland. keeled as in available specimens examined of either F. tigris or F. leo. The third metatarsal is as large and robust as in specimens of F. tigris, but the fifth metatarsal though fully as long as in the tiger is distinctly slenderer. A comparison of the foot elements from Cumberland Cave with those of Felis atroz from Rancho La Brea figured by Merriam and PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 53 Stock (1932, figs. 74, 92, 103, 106, 126, 130) shows the scapholunar, as- tragalus, caleaneum, and third metatarsal to be markedly similar to the corresponding elements in the California material but nearly one-third smaller. However, their dimensions are only slightly less than those given for a small individual in each case. The second metacarpal and fifth metatarsal in addition to being smaller are relatively slen- derer than in the Rancho La Brea form. A third metacarpal in the Port Kennedy collection, referred by Cope (1899, pp. 247-249) to F. inexpectata, compares favorably in length with the corresponding bone of a large puma but belongs to a foot much smaller than that indicated by the second metacarpal from Cumberland Cave. A fourth metacarpal from Conard Fissure bearing the number A.M.N.H. 11787, as does the type of FP. longicrus, is as robust as in F. tigris but markedly shorter, resembling that of a saber-tooth form. TaspLe 15.—Measurements (in millimeters) of foot bones of Felis cf. atrox See ee Measurement ve Ae Transverse width of scapholunar measured across medial portion of radial faceti ae s2s6552 44 ‘Greatest length of second metacarpal-_-_.------------------------------------------------------ 94 Greatest diameter of proximal end of second metacarpal-------------------------------------- 21 Greatest transverse diameter of proximal end of second metacarpal--------------------------- 15.2 Length of caleaneum between groove for achilles tendon and CubOlGliacet=se ee eee anne 99. 5 Width of caleaneum across sustentacular and ectal facets------------------------------------- 38 Length of astragalus between center of navicular facet and most remote portion of tibial facet on inner condyle..__.---=-------=--------------=--------------------=-=--<-=-=---------=--- 52 Greatest length of third metatarsal_-_--------------------------------------------------------- 113 Dorsoventral diameter of proximal end of third metatarsal__._.------------------------------- 29 Transverse diameter of proximal end of third metatarsal___.---------------------------------- 24 Greatest length of fifth metatarsal__---.------------------------------------------------------ 111.5 Greatest diameter of proximal end of fifth metatarsal _--_..------------------------------------ Vad Order RODENTIA Family SCIURIDAE MARMOTA MONAX (Linnaeus) In the collection from the Cumberland Cave deposit there are a number of specimens referable to a single species of woodchuck. They include seven nearly complete skulls, several lower jaws, and other bones of the skeleton. The better-preserved carry the National Museum catalog numbers 7997, 8127, 8157, 8158, and 8183. The cheek teeth in the cave specimens average somewhat less in size and the skulls average somewhat larger compared with a large series of the living species, but in these and all other differences noted they fall well within the range of individual variation of Marmota monaz. For this reason they are referred to the living species. 54 BULLETIN 171, UNITED STATES NATIONAL MUSEUM CITELLUS cf. TRIDECEMLINEATUS (Mitchill) A single specimen, an incomplete left ramus of a mandible, U.S.N.M. no. 12054, containing the incisor (broken at alveolus), M,, and Ma, represents the ground squirrels in the Cumberland Cave collection. This specimen is about the size of Citellus tridecemlineatus, to which it is provisionally referred. The two lower cheek teeth preserved are slightly less compressed anteroposteriorly relative to their width than is common in living specimens. ‘The absence of more complete material prohibits a satisfactory comparison with the Recent form. TAMIAS cf. STRIATUS (Linnaeus) A fragmentary lower jaw, U.S.N.M. no. 12367, included in the rodent material is recognized as belonging to a chipmunk. The lower cheek teeth are present though badly worn, and the incisor is broken away at the alveolar border. The specimen is not complete posterior to the last molar. The jaw is comparable in size and pro- portions to that of the living species, Tamas striatus. Nocharacters were observed that would serve to distinguish the fossil from the form now living in the same region. SCIURUS (TAMIASCIURUS) TENUIDENS Hay FIGuRE 30 The species Sciurus tenuidens was described by Dr. Hay (1920, pp. 104-105, pl. 4, fig. 20) on what he believed to be an upper incisor in a part of a skull from a cave deposit near Cavetown, Md. It was found upon removing the matrix that the specimen is a right lower jaw portion with the lower incisor. The cheek teeth were not pre- served. As an inferior jaw the specimen does not possess such marked characters distinguishing it from Sciurus hudsomcus. The portion of the jaw preserved is rather robust, and the incisor is rela- tively narrow transversely and of greater anteroposterior diameter than in S. hudsonicus. Also, the mental foramen appears to be placed lower on the side of the ramus than in the recent specimens examined. Sciurid material in the Cumberland Cave collection includes three skull portions and four mandibular rami. The form represented clearly belongs to the group Tamiasciurus but is apparently distinct from S. hudsonicus. The mandible is more robust and possesses a lower incisor of relatively great anteroposterior diameter and hence presumably represents the Cavetown form, S. tenwidens. On the basis of skull U.S.N.M. no. 8164 (fig. 30a and 6) and a nearly com- plete left ramus of mandible U.S.N.M. no. 12050 (fig. 30c) from Cumberland Cave, the species is redescribed. Specific characters.—Size about that of Sciurus fremonti, larger than the average of the S. hudsonicus group. Skull with notably broad frontal region and with short, deep muzzle. Viewed from the PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 55 side, occiput more elevated than in the living species, giving a less convex contour to the cranium. Anterior zygomatic region widely expanded, with its outer rim strongly turned downward and ex- tending forward well beyond the maxillary-premaxillary suture, thus giving an unusually broad surface for the attachment of the anterior portion of the masseter-lateralis. Jugal relatively deep, FIGURE 30.—Sciurus (Tamiasciurus) tenuidens Hay, skull (U.S.N.M. no. 8164), and mandible (U.S.N.M. no. 12050): a, Dorsal view of skull; 6, lateral view of mandible. X 14. Cumberland Cave Pleistocene, Maryland. about one-fourth deeper than the average in the living species. Bullae much as in the living members of the group but more inflated back- ward and outward, in posterior portion, giving a decidedly flatter appearance to the posterior wall of its inner face and to the outer face directly below the external auditory meatus. External audi- tory meatus smaller than in living species. Incisors more robust, with greater anteroposterior diameter than in the living species. Cheek-tooth row of equal length witb that of Sciurus fremonti, and of S. hudsonicus from the Fort Yukon district, but teeth slightly narrower. Lower jaw relatively deep, short, and massive, with broadly ex- panded angle. Anterior portion of masseteric fossa deeper and 56 BULLETIN 171, UNITED STATES NATIONAL MUSEUM broader and with more robust border ridge than in specimens of equivalent maturity. The red-squirrel material from the Conard Fissure was referred by Brown (1908, pp. 192-193) to Sciurus hudsonicus. Without a criti- cal comparison of actual material it seems that the Conard Fissure material represents a species close to Sciurus tenuidens, although certain features, as the slightly less robust incisors, the less extreme depth of the lower jaw angle, the apparently larger external auditory meatus, and the more convex longitudinal profile of the posterior portion of the cranium ally it perhaps more nearly to the living S. fremont of the Rocky Mountain region. TaBLE 16.—Measurements (in millimeters) of skull, mandible, and dentition of Sciurus tenuidens and S. hudsonicus Sciurus tenuidens Sciurus hudsonicus Measurement eae U.S.N.M. no. 9197. U.S.N.M.| U.S.N.M.|} no. 151654 (type) no. 8164 | no. 12054 (Div. Mamm.) Length, posterior margin of incisor alveolus to condyles-_-__|_-_.----__- AQND| aes 39.5 Depth of rostrum below nasofrontal suture__----...--.-.--]----------- AZjG Alea son ae 12 Width of skull across postorbital constriction_____........_]_-------_.- 45S eee. ee 13.5 Length of diastema between alveolus for I and P4...______|_-_--__--_- ADA NE ee eae ll Length of tooth series, P4 to M3, inclusive__..-..-...------|_...-.-___- Lael ee sae 78 Transverse: wid thiof M4. 222. 022 cos se es ee | eee ce ee asa (booed Pie Depth ofijawbelowsPas*=22seeeese ons. Seen en cose eee aa eee ee ee | eee 8.7 8.5 Length of diastema between alveolus for I and alveolus for Pea seae ncn eth ona sone ee ene seaen esac eoeee es eee 457 asoseeee ee 4.8 5.4 Length of tooth series from P, to M3 measured at alveolar DOR ers == 23225 ieee See te Soe ee ee ee 1 Sloe see seo 18.1 8.2 Anteroposterior diameter of lower incisor-_.-_------------- 8:9" |eenaccc ses 3.5 3.2 Transverse diameter of lower incisor_......----.----------- 1G) |ooccecwsces 1.5 1.2 ‘Transverse. width) ofiMy: <2. =t22 2 ct ae Seeeeie a. cae |e eee ee ere | eee ee as 2.2 2.2 1 Approximate. GLAUCOMYS species The flying squirrels are represented in the Cumberland Cave col- lection by a single specimen, a broken lower jaw, U.S.N.M. no. 7687, carrying three teeth, P, to M,. This fragment of jaw indicates a species larger than Glaucomys volans, not quite so large as G. yukonen- sis, and about the size of G. sabinus. The teeth combine characters of these three living species, with features of their own that suggest the specimen may represent a new species, but without more adequate material by which to determine possible ranges of variation it is preferred not to give it a definite assignment. The general characters noted in the Cumberland Cave specimen are as follows: Length of cheek teeth series P, to M.=5.3 mm; greatest width of M,=2 mm. (G. volans length of cheek teeth P, to M.=4.2; width of M,=1.6 mm.) P, relatively large, with deep talonid valley PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 57 that opens obliquely outward and backward through a notch that distinctly separates the hypoconid from the outward extension of the posterior rim of the basin. Paraconid, hypostylid, and entostylid of P, well developed. Basins of the molars broad and evenly concave. Space between protoconid and hypoconid relatively wide with very distinct valleys on each side of the well developed hypostyle. Inner rim of the cheek tooth crowns interrupted by a median, sharply defined notch. Family GEOMYIDAE PLESIOTHOMOMYS POTOMACENSIS Gidley and Gazin Ficure 31 Type.—Portion of left ramus of mandible (fig. 31), U.S.N.M. no, 8166, with all cheek teeth. Incisor not preserved beyond margin of alveolus. Generic characters—Anterior margin of symphysis of mandible very abrupt. Pit on lingual side and at base of ascending ramus shallow, as in Thomomys. Masseteric crest nearly straight and ex- tending high on mandible anteriorly. Lower molars oval in cross section and with enamel plates on anterior and posterior surfaces. Specific characters.—Size near that in average adult specimens of Thomomys fulvus. Mandible robust with deep symphysis. Incisor relatively wide with anterior surface smooth and rounded. Lower cheek teeth relatively narrow transversely. The pocket-gopher material in the Cumberland Cave collection referred to this species includes four other lower jaws, one of which, U.S.N.M. no. 12049, possesses a complete lower dentition. Comparison.—These specimens present several interesting charac- ters that seem to separate them clearly from both Thomomys and Geomys. The mandibles are remarkably deep and the symphysial portion is turned upward rather abruptly, causing the nearly straight anterior margin of the symphysis to form a sharp angle with the inferior margin of the ramus. Also, the noticeably straight masseteric crest rises anteriorly at a higher angle from the lower border of the jaw than in living forms, making the distance between the anterior extremity of the masseteric crest and the ventral limit of the symphysis relatively great. The dentition of these specimens is unique in combining characters of both Thomomys and Geomys. The incisor has its anterior surface rounded as in Thomomys but is relatively wide. P, has its enamel plates as in both Thomomys and Geomys, but the columns of this tooth are well rounded and the internal reentrant fold is compressed, suggesting Geomys. The external fold is more open. The molars are widely oval in shape, perhaps even less compressed antero- 13 Gidley and Gazin, 1933, pp. 354-356, fig. 7. 58 BULLETIN 171, UNITED STATES NATIONAL MUSEUM posteriorly than is common in Geomys. In Thomomys these teeth have their lingual portion tightly compressed and their buccal surface distinctly flattened and usually grooved longitudinally. However, OMG Wy) WY iy fi MW FIGURE 31.—Plesiothomomys potomacensis Gidley and Gazin: Left ramus of mandible, type specimen (U.8.N.M. no. 8166), lateral and occlusal views. 2. Cumberland Cave Pleistocene, Maryland. the molars in the fossil specimens have enamel developed on both anterior and posterior surfaces as in Thomomys. In Geomys only the posterior surface of these teeth has an enamel plate. TABLE 17.—Measurements (in millimeters) of mandible and dentition of Plesiotho- momys potomacensis and P. orientalis Plesiothomomys potomacensis Fe ere Measurement Pn at pee no. 8166 a= tt tuo zatads (type) (type) Superior dentition Length of diastema between alveoli of I and Py----.-------|----------------|---------------- 12 Anteroposterior diameter of upper incisor_-----------------|----------------|---------------- 11.8 Transverse diameter of upper incisor: - | -2-~---s-- 3 - ennnn| en nec ew ene e na | aan ae P4, anteroposterior diameter at occlusal surface---.---------|----------------|---------------- 22 P4, transverse diameter of anterior lobe_-_------------------|----------------|---------------- 1.6 P‘, transverse diameter of posterior lobe._--..--------------|----------------|---------------- 1.8 M!, anteroposterior diameter at occlusal surface.-.---------|----------------|---------------- aba Miitransverse diameter +2 02-2222 ate nanan a ease sens s secs | ewen tenn eae ane ae een 1.9 Mandible and inferior dentition Distance between lowest point on dorsal surface of mandi- ble anterior to Py and ventral extremity of symphysis-_--- 9 Sab Cassese Length of diastema between I and Py---------------------- 18.1 S..Gh| 22 assesses Length of lower cheek-tooth series measured at occlusal BUTIACO Ns ne eee eee eae nee nese nea aea eee 7 136 eos sensesoeee J anteroposterior: dlaMeterseses ses ae sane eee a eas | ae enna en 23 sees Sa soce Ve transverse diameter. 2-22 en. Se eee eae ae yal 2.2 Neonates Py, anteroposterior diameter at occlusal surface.----.------- 2.6 229 Seo 2o se eee P,, transverse diameter of anterior lobe_-_-.---------------- 1.3 Taio oees see es P4, transverse diameter of posterior lobe.------------------- 2 ae eh a Mi, anteroposterior diameter at occlusal surface--_---------- 1.4 155) || eee Mi, fransverseidiameter: -....-: =.) £2622. S222e- set se 2 Dr Dice. ee Saez ee Mi, anteroposterior diameter at occlusal surface__.--------- 1.5 136i [See eW seco Ms; transverse diameter. __-- 24-2022 ceaenconcese sees ae 1.9 2) aa ee See Ms, anteroposterior diameter at occlusal surface-_----------- 1.5 1.6) |be doses los Mis; transverse diameter=-----o-s.54-) ae ee leg, WSO) eee eaees wt 1 Approximate. PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 59 Thomomys orientalis Simpson (1928, pp. 6-7) from Saber-tooth Cave in Florida, although possessing characters that ally it to Tho- momys, apparently should be included in the genus Plesiothomomys. The upper incisor in the Florida specimen shows a single shallow eroove placed well toward the inner margin of the enamel, as in some modern species of Thomomys. Also, both P* and M? have the enamel complete across the posterior surface as in the western pocket gopher. However, the columns of these teeth are oval in cross section as in Geomys and the reentrant folds in P* are not so widely open as they are in Thomomys. Another interesting character observed in the type of P. orientalis is that the rostrum appears to be somewhat depressed anteriorly. This may be due in part to crushing, but the portion of the palate between the cheek teeth and incisors does not seem to ascend so steeply as in living species. P. orientalis is a distinctly smaller species with smaller teeth than P. potomacensis, corresponding more nearly in size to specimens of Thomomys fossor. Family CASTORIDAE CASTOR CANADENSIS Kuhl A single left upper molar in the Cumberland Cave collection, U.S.N.M. no. 12056, may be so exactly matched in specimens of the living species Castor canadensis that it must for the present be referred to this species. Height of crown (apparently not much worn) 23 mm; anteroposterior length, triturating surface, 6.5 mm; actual antero- posterior diameter 6 mm, width 6.5 mm. Family CRICETIDAE PEROMYSCUS cf. LEUCOPUS (Rafinesque) The deer mice are represented in the cave collection by eight frag- mentary mandibles with incisors. Seven of the specimens have one or more cheek teeth each. A comparison of this material with jaws of Peromyscus leucopus from the Recent fauna of Maryland failed to reveal any characters of diagnostic importance separating the fossil from the living form. Although more complete material including skulls may show important specific differences, this is not suggested in the mandibles. Several modern species of Peromyscus, as P. mani- culatus and P. leucopus, have remarkable geographic distribution, and it seems not at all unlikely that their geologic range through Quater- nary time may have been equally extensive. NEOTOMA MAGISTER Baird # FicurE 32 Three wood-rat mandibles, U.S.N.M. nos. 12033 (fig. 32), 12034, and 12035, in the collection apparently represent the large Neotoma 13 Baird, 1857, p. 498. 16541—38——5 60 BULLETIN 171, UNITED STATES NATIONAL MUSEUM magister. The specimens compare favorably in size with the cotype material from Carlisle Cave, Pa., being somewhat larger and deeper jawed than in an average living specimen of N. pennsylvanica.* The Cumberland Cave specimens are characterized by a relatively long mandibular symphysis, robust incisors, and cheek teeth with postero- external reentrant folds widely U-shaped and deeply pocketed at base. FIGURE 32.—Neotoma magister Baird: Left ramus of mandible (U.S.N.M. no. 12033), lateral and occlusal views. 2. Cumberland Cave Pleistocene, Maryland. The Cumberland form is readily distinguished from N. floridana by its greater size, larger incisors, and relatively longer diastema between incisor and cheek teeth. N. ozarkensis Brown (1908, pp. 196-197) is apparently somewhat smaller than the Cumberland Cave wood rat, presumably nearer in size to N. pennsylvanica. TaBLE 18.—Measurements (in millimeters) of mandible and inferior dentition of Neotoma magister U.S.N.M. | U.S.N.M. Measurement no. 12033 | no. 12034 Length of lower molar series measured at occlusal surface__-_------.-------------- 8.8 9.4 Length of lower molar series measured along alveoli-_----------------------------- 9.7 10 Mi anteroposterior length a3: 2222. = So ee ee ee 3.4 Sov My itransverseswidth-e= = 225 2 2 he ee ee se ee eee senaee eee aee 1.9 2.1 Ms, ‘anteroposterior lensths soo 8 ea oa ee See eene ene oeasecee 2.8 3 Mo, transverse’width. : 2i0i1222252 622228 oe os case csastaaeccteeosee 2.2 a M3; anteroposterior length =.----2--<2=-< 222s sss3— a Rot serene dee ee oscetaessas 252) 2.3 Msitransverse: width. 22.50. 222 es es ee ence en cseeseeee 1.9 1.9 Depth ofiramus'at junction of; May and "Migs ~ o= = es ee esa se neon ean 5.6 5.8 PARAHODOMYS SPELAEUS Gidley and Gazin '5 FicurREs 33, 34 Type.—Portion of the right ramus of a mandible (young individual), U.S.N.M. no. 12040 (fig. 33), with all cheek teeth and part of alveolus for incisor. Generic characters —Cheek teeth in general structure as in Hodomys, but with reentrant valleys, especially the external ones, shallower and 44 The possibility of N. pennsylvanica Stone being asynonym of N. magister Baird was noted by Rhoads, 1894, pp. 213-221. 18 Gidley and Gazin, 1933, pp. 356-357, figs. 8, 9. PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 61 directed nearly at right angles to the line of the tooth row. [They are directed more obliquely forward in Hodomys.] Anteroexternal valley of M, and M; very shallow and absent in M;. [In Hodomys M; has a shallow but well-defined anterior reentrant in young specimens.] Mz, with a well-defined posterointernal reentrant valley. [This fold is absent in Hodomys.] As seen in mandibles U.S.N.M. nos. 12037 (fig. 34) and 12041, the following additional characters appear as of generic importance: FIGURE 33.—Parahodomys spelaeus Gidley and Gazin: Right ramus of mandible, type specimen (U.S.N.M, no. 12040), lateral and occlusal views. <2. Cumberland Cave Pleistocene, Maryland. Symphysis relatively deep. Lower branch of masseteric ridge very prominent and sharply defined. Condyle relatively large and de- pressed so that condylar bar is directed backward to a much greater extent than in Hodomys and the notch between it and the coronoid process is much more widely open. Specific characters—Size about that of Hodomys alleni. Incisors relatively robust, causing a more prominent swelling of the lower FicuRE 34.—Parahodomys epelaeus Gidley and Gazin: Right ramus of mandible (U.S.N.M. no. 12037), lateral and occlusal views. X 2. Cumberland Cave Pleistocene, Maryland. border of the ramus than in H. allent. Diastema between incisor and cheek teeth long. Posteroexternal reentrant folds in all molars shallow and pocketed at base, wearing at occlusal surface in wide open U-shaped pattern from summit to base of crown. Postero- internal reentrant fold well developed in M3, extending somewhat below middle of crown but disappearing in old age. The material referred to this form includes seven lower jaws in addition to the type, representing individuals in several stages of maturity. Comparison.—The genus here described is known only from lower jaws but fortunately these reveal diagnostic characters that clearly distinguish it from the nearest related living genera, Neotoma and 62 BULLETIN 171, UNITED STATES NATIONAL MUSEUM Hodomys. Of these the most readily determinable characters are to be found in the last lower cheek tooth. In old individuals, showing well-worn teeth, this tooth, in fossil as well as in living genera, has but two reentrant angles, one on the inner, the other on the outer face of the crown. In Neotoma these reentrant valleys are oppositely placed and tend to divide the crown surface into two subequal areas. In Hodomys and Parahodomys the reentrants alternate, the inner being well in advance of the outer one, tending to form a more or less S- shaped enamel pattern on the triturating surface. However, the latter genera are readily distinguished from each other by the fact that in Parahodomys the reentrant valleys are directed almost at right angles to the longitudinal plane of the tooth row, while in Hodomys they slant well forward. In young individuals with slightly worn teeth, additional distinctions are apparent in that there is present a third reentrant fold in this posterior molar in all three genera, but this fold is peculiarly situated in the fossil form. In Parahodomys it occurs on the posterior internal lobe opposite the large external fold and extends downward to a point slightly below the middle of the crown. In Neotoma and Hodomys the third reentrant fold is on the anterior lobe of the outer side. It occurs only very near the summit of the tooth in Neotoma, but it extends down to about the middle of the tooth crown in Hodomys. TapLE 19.—Measurements (in millimeters) of mandible and inferior dentition of Parahodomys spelacus and Hodomys alleni me Parahodomys spelaeus Hodomys allenit a USMM. | ews. | Won no. 12040 Yona no. 91525 (type) no. 12037 | (Biol. Surv.) See eee ee Total length of jaw, condyle to incisor border-.------------- eee ae ee 30 28.7 Depth of symphysis-.--------------------------------------]--------------- 5.4 4 Depth of ramus at point between M2 and Ms3--------------|--------------- 7.3 6.5 Height of condyle above level of cheek tooth row-----------|--------------- 7 10 Length of cheek tooth row measured at alveolar borders---- 9.5 9.5 10.8 Mi, anteroposterior diameter---...------------------------- SON Eee eee 3.6 M,, transverse diameter ----------------------------------- Oe ee eae | 2.3 Mz, anteroposterior diameter ---...------------------------- 2.8 3 3.5 M:,, transverse diameter ---.-------------------------------- 2.1 2.4 2.5 M3, anteroposterior diameter - --.-------------------------- 2.4 2.6 2.9 Ms, transverse diameter- --------.------------------------- 2 2 2.1 eee eee — SYNAPTOMYS cf. COOPERI Baird The anterior portion of a single lower jaw, U.S.N.M. no. 7774, in the Cumberland Cave collection, carrying the incisor and first molar, represents a species of Synaptomys that corresponds in size and is not otherwise clearly distinguishable from S. coopert. In the many speci- mens of the living species of this genus examined the triangles of the PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 63 molars are more tightly closed than in the fossil specimen. This is the only difference observed, but if it could be verified in a large num- ber of specimens it might constitute a species distinction. SYNAPTOMYS (MICTOMYS) species The subgenus Mictomys is also represented in the Cumberland Cave collection by a single specimen, a broken lower jaw, U.S.N.M. no. 7773, with complete dentition. No characters were observed by which the fossil form could be distinguished from the living species, S. (M.) borealis. This species is currently recognized to contain all the races within the subgenus Mictomys. However, all the lemming mice of this group are at present Canadian and Alaskan in distribu- tion (Howell, 1927, p. 19), extending down into northern Washington and northern New Hampshire in the United States. Although the fossil appears to demonstrate the presence of this modern boreal spe- cies in the Pleistocene of Maryland, its occurrence so far outside of the present distribution suggests that a distinct species may be represented. Additional importance is to be attached to this occurrence in furnishing the first Pleistocene record of Mictomys (Gidley, 1913b, p. 96). MICROTUS (or PITYMYS?) cf. INVOLUTUS (Cope) FiGurE 35 The voles are very sparsely represented in the Cumberland Cave collection. Seven fragmentary specimens, one a part of a palate with the two anterior cheek teeth on the right side, the others broken mandibles, comprise the lot. The material is clearly microtine in character but is apparently distinct from North American living species. The root portion of the lower incisor in the Cumberland Cave form passes between the lower portions of the penultimate and ultimate molars as in voles, and the cheek teeth are without roots in those specimens where this character can be observed. However, the pattern of the first lower cheek tooth is somewhat simpler than in the living pine-mice of the region. This tooth, presumably M,, is characterized by a posterior loop followed forward by three closed triangles, a pair of confluent triangles opening into an anterior loop. In these characters the fossil jaw resembles Pitymys but unlike the form living in the same region at the present time, P. pinetorum, the anterior loop of M;, in three of the six jaws is without reentrant angles. Two specimens have shallow reentrant angles on both inner and outer surfaces, nearly as in P. pinetorum, and one has a shallow groove on the antero-internal surface of the anterior loop (fig. 35). Furthermore, the posterior surfaces of the triangles in the lower cheek teeth are slightly more convex, andisola- tion of the triangles is not so marked as in specimens of P. pinetorum. 1¢ Cope, 1871, pp. 89-90, fig. 16. 64 BULLETIN 171, UNITED STATES NATIONAL MUSEUM Also M3; is somewhat more robust than is common in the Recent species. Microtus pennsylvanicus, the typical microtine form living in Mary- land today, exhibits a more complex pattern in the first lower tooth. FicureE 35.—Microtus (or Pitymys?) cf. involutus (Cope): Right ramus of mandible (U.S.N.M. no. 12386), lateral and occlusal views (pattern of M3 taken from mandible U.S.N.M. no. 7772). 3. Cumberland Cave Pleistocene, Maryland. Presumably M. pennsylvanicus is to be regarded as possessing a more advanced type of dentition. TaBLE 20.—Measurements (in millimeters) of dentition of Microtus(?) cf. involutus and M. dideltus Microtus(2) ef. involutus ee Measurement T8s#| 0.8,°}Us8s (U8. | AcE ALM) N.M.|N.M.|N.M.|N.M.| N.H. | N.H. Ol no. no. no. no. no. no. 8694 7686 | 12602 | 7772 | 12368 | 8699a | 8695 (type) Superior dentition T,anteroposterior diameter.ss-.cas-2ees eer eeeeoces oe 12 Ssccce alte sie Se be a Eee oe | ban oe eeeeroe i; transverse! diameter: =222-. = 2-2 252-522226 2. -ses2—52 Le [eect ot boos te Ss Sets ot sete ee ao ae M!, anteroposterior diameter !__.-.....-.------------ 2. 2h Soc arene |ot eee | Se owes 24 jot oe M1. transverse Giameters.s.2- 2 .2s222525-2-<-cesecces Veal eos elesce kan lan oese al oence se 143)'|Ss22255 M,, anteroposterior diameter__..........------------ D6) sees sce sok ces eeeececb ee 18s ota M3" transverse: diameters. 2- sos ae aneeeoassasoeees Li: Abeseeesel bese dsehee sccu laste =e L2hfsssone - M3 anteroposterior diameter-s-c2o2sco2- ee see ee sas | sao cee joc ae cre Coeeeee | peeeees loaeenes LO icone = Ms; transverse diamoters22222.2csh soc acs cocoa nen c= lobes ces) sense eth cote aes ae ae |eoeeees P21 ccee, 2 Inferior dentition Z,janteroposterior diameters 222 oe. ses eee ee een eee 1,2 1,2 1330 tosdsccs|souuc ols I, transverse diameter. 2222-22 25-2 ses sk 2 ese e ee es se|s2ess=5 0.9 1 1 Lee eesee 91.5 My; anteroposterior diameter22-.1—- =. 22. sce coe anne |e ates 2.6 2.5 at 26 lssese 3 Mia transverse: diameters22-2 <2 ose ooo oe eee as |oe ee 1d 1 1 1.4 jAcescce 1.2 Ms; anteroposterior diameter-s2--2-saeenscanseessen|eeaaee 1.5 1.5 1G 48020 |e eee eeeeees Ms, transverse diameteris:2 222-28 cece een wenn eet 0.9 0.9 1 Jabatss Clee Sete ee M3; anteroposterior diamoter-c: 224. co- sas aa=- sos |o=ee see 1.5 1,43) css-5 173. Joo 33S] sees 3 IMizs trans Verse dismMetebesss=se sss ees ae eee eee anne peeee 0.8 OT Rencece ON2|Poaeceo|Sassees Length of lower cheek-tooth series measured at occlusal jsurface...% .< 22. 4-<..s22eeeo Seed ens on |Saeneee Bad Blleccenns 95.6) | scsc-es -tcesse 1 Anteroposterior dimensions taken at occlusal surface. 21 Dimension as given by Cope but not verified. 4A pproximate. The Cumberland Cave form corresponds closely in size to the type of Microtus? involutus (Cope) from the Pleistocene of Pennsylvania. PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 65 The enamel pattern of the lower cheek teeth is markedly similar in the two forms. It is important to note that the anterior triangles in the first lower cheek tooth of the type are confluent as in Pitymys, not as in Cope’s illustration. This error in the drawing tends to exagger- ate the appearance of the anterior loop, which in reality is not greatly different from that in some of our specimens. Cope noted the re- semblance between his form and P. pinetorum, but apparently the Pleistocene form is distinct from the living species in exhibiting a slightly simpler enamel pattern. The teeth in Microtus(?) speothen (Cope, 1871, pp. 87-88, fig. 13) and M.(?) dideltus (Cope, 1871, p. 89, fig. 15) are distinctly larger than in the Cumberland Cave form. In M.(?) speothen the enamel pattern in the first lower cheek tooth is of a different type than that in Pitymys. This tooth in M.(?) dideltus, while showing confluent anterior triangles as in Pitymys, has an anterior loop that is narrow and elongate, and the portion of the tooth posterior to the loop is relatively wide. As noted by G. S. Miller, Jr. (1896, p. 59), the anterior lower cheek tooth of Pitymys is not different from that in Pedomys, or Microtus ochrogaster, living in the Mississippi Valley. Hence, reference of the Cumberland fossil to the genus Pitymys in the absence of satisfactory skull material cannot be made with certainty. ONDATRA cf. ANNECTENS (Brown) FIGURE 36 A very small species of Ondatra is represented in the Cumberland Cave material by a broken mandibular ramus (fig. 36), U.S.N.M. no. 12044, containing all the teeth. In size this specimen agrees closely with the type of the Conard Fissure species, O. annectens, being only slightly smaller and possessing a somewhat narrower incisor. It agrees with O. annectens in most other important details observed and is therefore provisionally referred to this species. It was suggested by Hollister (1911, pp. 33-34) that Brown had wrongly interpreted some of the characters described. Brown con- sidered the Conard Fissure species as ‘‘an intermediate type connecting Fiber (=Ondatra) with Microtus through the subgenus Neofiber.”” The type of 0. annectens as observed by Hollister indicates “a species close to existing forms in everything except size’, and he stated that the “anterior loop of first lower molar is more deeply cut by the reentrant angles than in any specimen of an existing species examined.” If these anterior reentrants are to be considered within the loop, as implied by Hollister, then the number of reentrants in M, posterior to the loop on each side appears to be one less than in 0. zibethica, approaching the condition seen in Neofiber alleni. In either case the 17 Brown, 1908, p. 197. 66 BULLETIN 171, UNITED STATES NATIONAL MUSEUM columnar portion anterior to the fifth reentrant, counting forward along the inner side of the tooth, is not so well developed in O. annec- tens as in O. zibethica. Advanced wear in M, of O. zibethica may show this column or loop to be without reentrants although well developed. In the description of O. annectens Brown called attention to the markedly reduced condition of the anteroexternal column of M3. This column is not more reduced in the Cumberland Cave specimen than in certain specimens of O. zibethica. In Neofiber this column is greatly reduced. In both the Cumberland Cave and Conard Fissure jaws the surfaces for the attachment of the masseter muscles are nearly alike but differ slightly from those of the living species of Neofiber and Ondatra. CRE ND FIGURE 36.—Ondatra cf. annectens (Brown): Left ramus of mandible (U.S.N.M. no. 12044), lateral and occlusal views. XX 2. Cumberland Cave Pleistocene, Maryland. The masseteric crest in general is bulging and sharply keeled, and there is a well-defined groove marking the posterior margin of the area for attachment of the anterior masseter medialis. The distinguishing feature, however, lies in the position of this sulcus, which in O. annec- tens has its origin much farther back from the anterior end of the mas- seteric ridge from where it curves sharply upward, leaving a definitely wider margin between it and the anterior border of the ascending ramus. It is further noted in the Conard Fissure jaw that the shaft leading to the condyle carries a fairly well developed sulcus on its anterior face nearly as in Ondatra zibethica. This sulcus is incipient or absent in Neofiber. TABLE 21.—Measurements (in millimeters) of inferior dentition of Ondatra ef. annectens A.M.N.H. Measurement ue Perk no. 12424 ee (type) Length of cheek-tooth series measured at alveolar margins_--_------------------- 111.5 12 Length of cheek-tooth series measured along occlusal surface____.---------------- 10.5 10.9 Mi, anteroposterior length at occlusal surface__......---.------------------------ 5.1 5.4 Mistransverse: width=. \- 2... $2 ee ee ee ek oe ee eee 2.1 2.2 M2, anteroposterior length at occlusal surface__...--...----.--------------------- 27 2.8 Mostransverse: widthisscs. 23. ote re ne en eR i ee ae 2 2 Ms, anteroposterior length at occlusal surface._-..---.--------------------------- ONT a7 Mis} transversetwidth:_<¢ Fs. - 2 seh be, re A eg ee eles peek eS 1.6 1.6 1 Approximate. PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 67 O. annectens seems to stand as a well-defined species of Ondatra but is presumably less advanced than the living species. Neofiber may be considered as still less advanced in the development of the pattern reached in M,, although more advanced in degree of hypsodonty. Family ZAPODIDAE ZAPUS species The genus Zapus is represented in the Cumberland Cave material by part of the right ramus of a mandible, U.S.N.M. no. 7777, carrying all the cheek teeth. This jaw represents an old individual in which the teeth are considerably worn. They nevertheless present the characters peculiar to the genus to which the specimen is referred and show certain peculiarities that prevent assigning the form to any known species. The material, however, is not sufficiently complete to warrant establishing a new species. The tooth crowns of M, and M; are relatively shorter and the an- terior main lobe of M, is relatively wide as compared with those examined of a large series of specimens of the living species. NAPAEOZAPUS cf. INSIGNIS (Miller) A second lower jaw, U.S.N.M. no. 12366, of a jumping mouse in the collection is recognized as belonging to the woodland form Napaeo- zapus. This jaw fragment is from the left side and has the incisor and last two molars preserved. M, is represented by the alveoli and root portions. The jaw is distinctly larger and has larger teeth than the specimen referred to Zapus. Moreover, the two molar teeth show a tightly folded, complex pattern on the occlusal surface, which re- sembles Napaeozapus more closely than Zapus. The fossil compares favorably in proportions with specimens of the living species, N. insignis, differing in having a slightly more robust incisor, which appears to be somewhat less curved than is common in the modern form. Also, the alveoli for M, suggest that this tooth may have been a little longer anteroposteriorly than in N. insignis. Family ERETHIZONTIDAE ERETHIZON cf. DORSATUM (Linnaeus) Figure 37 The porcupine is represented in the Cumberland Cave collection by about 15 determinable specimens, including the greater part of 3 skulls and 12 broken lower jaws. In North America, according to Miller’s list of Recent mammals (Miller, 1924), there are two living species of Hrethizon now recog- nized: E. dorsatum and E. epixanthum. 'The former comprises two subspecies and the latter five subspecies. Apparently the distinction between the two full species is based solely on color differences in the 68 BULLETIN 171, UNITED STATES NATIONAL MUSEUM pelage. In an examination of a series of Recent skulls and jaws in the collections of the U. S. National Museum and U. S. Biological Survey, it was found that they show a wide range of individual and sex variations and that it was apparently not feasible to formulate a series of definite characters that would satisfactorily separate the two species. The specimens from Cumberland Cave likewise show considerable variation but apparently not exceeding in any important details the limits of individual variation seen in modern specimens. FIGURE 37.—Erethizon cf. dorsatum (Linnaeus): a, Skull (U.S.N.M. 7996), lateral view; 6, mandible (U.S. N.M. no. 7672), lateral and occlusal views. Two-thirds natural size. Cumberland Cave Pleistocene, Maryland. A remarkably large skull (fig. 37a), U.S.N.M. no. 7996, in the cave collection is characterized by a noticeably broad frontal region, wide palate, and particularly broad incisors. The incisors appear to be slightly wider than in the Recent specimens examined, although other proportions characterizing this skull seem to be duplicated in some of the larger of the modern specimens. Order LAGOMORPHA Family OCHOTONIDAE OCHOTONA species Six fragmentary mandibles and three maxillary portions, all carry- ing teeth, are recognized as belonging to a species of Ochotona. With this limited material no important differences can be cited separating PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE 69 it from the living species of North America, nor can it be assigned to any modern form with certainty. The form corresponds favorably in size with specimens of both O. princeps and O. schisticeps. Although the dentition is closely the same as in these Recent species, the styles or columns on the inner and outer surfaces of the lower molariform teeth in a couple of the fossil mandibles are slightly closer together than is common in North American Qchotona. This condition sharpens or narrows somewhat the inner and outer reentrant angles. The presence of Ochotona in the fauna is significant inasmuch as the North American species are found living only in the western and northwestern part of the country. The present occurrence shows the pika to have been present in the Appalachian region in Pleistocene time. The fossil record of Ochotona is very scanty. A form described by Cope (1871, pp. 93-94) as Praotherium palatinum from the Port Kennedy deposit may possibly be Ochotona. The Cumberland Cave specimens, however, do not conform to the description given by Cope. The upper molariform teeth of the Maryland Ochotona show a pro- nounced external groove as in the living species and an enamel ridge divides the occlusal surface of each tooth. Furthermore, the first, second, and last teeth of the cheek series are structurally distinct from one another and from the two intermediate teeth, as in living pikas. According to Cope the four upper teeth preserved in the tooth row of P. palatinum are structurally alike; also, their external surfaces are not markedly grooved and the occlusal surface of each does not show a distinct median ridge. Family LEPORIDAE LEPUS cf. AMERICANUS Erxleben Hares are well represented in the cave collection. At least 10 skull portions and 19 lower jaws are included in the material. A noticeable range in size and proportions of skulls, jaws, and teeth exists, though apparently within the limits of individual variation. No persistent characters of importance were noted distinguishing the fossil from the living hare, L. americanus. However, it is possible that some of the smaller, more incomplete lower jaws may represent Sylvilagus, but none of the skull portions shows the distinct interparietal and postor- bital processes formed as in Sylvilagus. 70 BULLETIN 171, UNITED STATES NATIONAL MUSEUM Order PROBOSCIDEA Family MASTODONTIDAE MAMMUT cf. AMERICANUM (Kerr) A mastodon is represented in the Cumberland Cave collection by four juvenile teeth and a tibia without epiphyses. The teeth are apparently the second, third, and fourth left and the third right of the lower-jaw series. There is nothing to distinguish these scanty remains from the typical Mammut americanum. Order PERISSODACTYLA Family EQUIDAE EQUUS species The genus Hquus is represented in the Cumberland Cave collection by only three teeth and three toe bones. The teeth, an upper and a lower of one individual and an upper of a second individual, are those of juveniles; they represent the milk dentition. The toe bones are two first phalanges and a metacarpal, representing adult individuals. These specimens belong to one of the large species of eastern horse, as E. complicatus or FE. pectinatus, but on this incomplete material an attempt to determine the exact species represented would be no more than guess work, based on size alone. One peculiarity of the foot bones may be mentioned. Both the metacarpal and phalanges are relatively wide for their length and depth as compared with the western species of Pleistocene horses, suggesting a species with short stocky limbs. Family TAPIRIDAE TAPIRUS species Figure 38 Two last lower molars (see fig. 38) of the right side belonging to adult individuals, and a fragment of a left upper premolar of a third and much younger individual, represent the tapirs in the Cumberland Cave collection. These teeth indicate a species somewhat larger than the living tapirs of South and Central America and the Florida Pleistocene tapir, 7’. veroensis Sellards, but not quite so large as T. haysii Leidy. Also they are of narrower proportions than the corresponding tooth in the Port Kennedy specimen, which has been assigned to 7. haysii. In astudy of lower teeth of a series of the living species of tapirs, the proportions PLATE 10 BULLETIN 171 U. S. NATIONAL MUSEUM ‘wnosnyy [PUOHEN *S “f) Ul poUqiyXy ‘sAVD puvjIoquINny Woy (Ad] pry SISUAPUDILIQUNI SN uUosKy] q) Areooad Jo uojaays poyunojy PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE Te of length to width seem to run fairly constant; hence the extreme narrowness of the molars from the Cumberland Cave suggest they FIGURE 38.—Tapirus sp.: Last lower molar (U.S.N.M. no. 7667), lateral and occlusal views. Natural size. Cumberland Cave Pleistocene, Maryland. may represent an undescribed species. But a more complete dental: series is necessary to confirm such a conclusion certainly. Order ARTIODACTYLA Family TAYASSUIDAE PLATYGONUS CUMBERLANDENSIS Gidley Fiaures 39-45; Puatre 10 Synonym: Platygonus intermedius GIDLEY. The type of this species is a nearly complete skull (figs. 39-41), U.S.N.M. no. 8146, including all the cheek teeth but without canines or incisors, and the symphysial portion and left ramus of mandible with inferior cheek teeth and both canines. A nearly complete skele- ton of an adult female, U.S.N.M. no. 8200, was designated as a paratype. The peccary skeleton (see pl. 10) has been mounted and is on ex- hibition with the bear and wolverine skeletons also from the cave. The skull and mandible belonging to this specimen are slightly crushed but required no restoration. Nearly all the elements of the vertebral column are preserved; only the tips of the spines and processes needed. modeling. Many of the ribs were preserved and much of the sternum. The pectoral and pelvic girdles were badly fractured and required some patching. The fore limbs and feet are complete except for the distal portion of the right ulna and one left and two right carpals, which are represented in plaster. The hind limbs are also nearly 18 Gidley, 1920b, pp. 659-673. 2 BULLETIN 171, UNITED STATES NATIONAL MUSEUM entire; only the left required restoration. All the parts of the hind feet are represented, but in this case one or two elements have been substituted from the general collection. Peccaries were one of the commonest forms found in the cave, ex- ceeding even the black bears in quantity of material. There are 22 skulls more than half complete, 12 of which are nearly entire. Seven- teen mandibles including the greater part of both rami are preserved. In addition to the skeleton described above, there is also an articulated hind foot (fig. 45) and a large number of isolated vertebrae and limb and foot bones. The following characters were given by Gidley as distinguishing P. cumberlandensis from other Pleistocene species of Platygonus: A large species, nearly equalling P. vetus in size. Length of cheek-tooth series, type (male) 94 mm.; paratype (female), 87 mm. Differs also from P. vetus in having a greater relative length of the molars in animals of corresponding age, and in the possession in the upper molars of well-developed intermediary cusps and lophs, which connect at their bases the two principal cross lophs, as in P. compressus and P. leptorhinus. It differs from both these last named species in the much larger size of the skull, in which there is a relatively higher and more backwardly produced inion and a more strongly developed expansion of the zygoma, the latter being about one and one-third times the vertical diameter of the orbit in the females, and two or more times this diameter in the males. Apparently the most distinctive features of this form are the elon- gate skull, the marked backward projection of the inion accompanied by a decided backward-sloping occiput, and the unusual extension of the zygoma downward and forward from the orbit. The develop- ment of the zygoma varies considerably with the age of the individual and between sexes, but it seems to have progressed farther in this species than in others as known. A second species described by Gidley as Platygonus intermedius (figs. 42-44) does not appear to possess sufficiently well defined characters to justify distinct recognition. The relative sizes of teeth and skull and the length of the diastema between canine and second premolar are not persistent enough throughout the series of specimens to war- rant this specific division. A comparison of the limb and foot material of P. cumberlandensis with that described by Leidy (1889a) as belonging to P. compressus and that for P. leptorhinus by Williston (1894) disclosed important characters in the Cumberland Cave form. P. cumberlandensis pos- sesses longer and heavier limb bones in keeping with the larger and more elongate skull in this form than in either P. compressus or P. lep- torhinus. This may be noticed even in the relatively slightly built female skeleton. On the other hand, the foot bones, particularly the metacarpals and metatarsals, though usually as large and in some instances larger than in the two earlier described species, are relatively short in proportion to the limb elements. This is clearly illustrated in the proportions of the fore and hind limbs of the articulated skeleton. 73 PLEISTOCENE VERTEBRATES FROM CUMBERLAND CAVE “PUB[AIRIA ‘BTED0ZSIOTq OABO PUBLIOQMING “9ZIS [BINICU J[VY-PUD “META [E199 E] “(QFIQ (OU “JA N'S'Q) Touroeds ed.Aq ‘TING :AO[PTD sisuapunpiaquinds snUobhjojq—'6E HANOI AL _ CN Al \) é 7% SAAN at aK RMON SN << LRN ae ae | BULLETIN 171, UNITED STATES NATIONAL MUSEUM 74 ‘eZ]S [RANJLU J[VY-OUQ “MOTA [VIZWOA “(9FTg “OU “JA'N“S"A) VompOads od44 [NYG :AO[PIH s1suapunpsaquns snuobhyn}g— py AAAOTA : a SS Ss f WIINN OO y YAWN 7 SN - a cs aT a 1 ei x t > QS - im SY ee Up 4) Bee PP Nat 4 ie tS 4 = re NY a ws oe he NP RG EAR, 2 NGS jy ES eS ye SY Ama Ra ee ee a Me ee ee ee ea ASS ~~ I y = ee . mh ii fi aw Ui ] Wy i Ee el YEA. ss eee : ; : p21 Lik SS NOS ANAVN WR ppp Dy << S IN i J fe Upp all SSeS ee Ws 1 VR} LNG sa \, hee 1) 75 ‘On[s [eINJeU J[ey-etQ “AOSTA [eSIOp “(9FT8 “OU “W'N'S’N) Uoumoods od 44 “[INyS :AQ[PID s[suapuDpsaqund snuobhzojgJ— Th LANOIY Yj yi, 4 Z YY i) WY : Yj by 7 Uf yy YY M AN Yyy a SSS SS LLL; SA Se ‘ WS AW \ LY KAU Wy lod, pil | Va : YY ee » i Za Se < yy SS _LIzvUJJJfyyv EON» S\N fee as Se y Yj Z, Ys ty tj Gy TG \ oe WN HOLY LOAVY Z - Ss, Se EaF Ze i . ’ ' . DSSS Sains : “NN ae