FIELDIANA Geology Published by Field Museum of Natural History New Series, No. 7 REVIEW OF THE HATHLYACYNINAE, AN EXTINCT SUBFAMILY OF SOUTH AMERICAN "DOG-LIKE" MARSUPIALS LARRY G. MARSHALL MAY 5 m\ April 30, 1981 Publication 1318 REVIEW OF THE HATHLYACYNINAE, AN EXTINCT SUBFAMILY OF SOUTH AMERICAN "DOG-LIKE" MARSUPIALS FIELDIANA Geology Published by Field Museum of Natural History New Series, No. 7 REVIEW OF THE HATHLYACYNINAE, AN EXTINCT SUBFAMILY OF SOUTH AMERICAN "DOG-LIKE" MARSUPIALS LARRY G. MARSHALL Assistant Curator of Fossil Mammals Department of Geology Field Museum of Natural History Accepted for publication July 11, 1980 April 30, 1981 Publication 1318 Library of Congress Catalog Card No.: 80-69140 ISSN 0096-2651 PRINTED IN THE UNITED STATES OF AMERICA CONTENTS List of Illustrations vi List of Tables viii Abstract 1 Introduction 1 Abbreviations 2 Acknowledgments 2 Systematics 5 Superfamily Borhyaenoidea 5 Family Borhyaenidae 5 Subfamily Hathlyacyninae 5 Patene 6 P. simpsoni 7 P. coluapiensis 13 Procladosictis 16 P. anomala 16 P. erecta (nomen xnnum) 18 Pseudocladosictis 18 P. determinabile (nomen vanum) 18 Pseudonotictis gen. nov. 19 P. pusillus 19 Notictis 24 N. ortizi 24 Perathereutes 26 P. pungens 26 Borhyaenidium 30 B. musteloides 31 B. nggsi sp. nov 37 Sipalocyon 40 S. externa 42 S. gracilis 44 S. obusta 60 Notocynus 63 N. hermosicus 63 Notogale 65 N. mitis 65 N. tenuis (nomen vanum) 72 Cladosictis 73 C. centralis 73 C. patagonica 79 Chasicostylus 95 C. castroi 95 Anatherium 99 A. herrerae sp. nov 103 A. defossus 103 Hathlyacyninae, Indeterminate 108 Summary of Evolution of Hathlyacyninae 108 Literature Cited 117 LIST OF ILLUSTRATIONS 1. Map of South America showing some vertebrate fossil localities 3 2. Map of southern tip of South America showing vertebrate fossil localities .... 4 3. Patene simpsoni Paula Couto, 1952a (Riochican) 7 4. Patene simpsoni Paula Couto, 1952a (Riochican) 8 5. Patene simpsoni Paula Couto, 1952a (Riochican) 9 6. Patene simpsoni Paula Couto, 1952a (Riochican) 10 7. Patene simpsoni Paula Couto, 1952a (Riochican) 11 8. Patene simpsoni Paula Couto, 1952a (Riochican) 13 9. Patene coluapiensis Simpson, 1935 (Casamayoran) 14 10. Patene coluapiensis Simpson, 1935 (Casamayoran) 15 11. Procladosictis anomala Ameghino, 1902b (Mustersan) 16 12. Procladosictis anomala Ameghino, 1902b (Mustersan) 17 13. Pseudonotictis pusillus (Ameghino, 1891c) (Santacrucian) 20 14. Pseudonotictis pusillus (Ameghino, 1891c) (Santacrucian) 22 15. Pseudonotictis pusillus (Ameghino, 1891c) (Santacrucian) 23 16. Pseudonotictis pusillus (Ameghino, 1891c) (Santacrucian) 24 17. Notictis ortizi Ameghino, 1889 (Huayquerian) 25 18. Notictis ortizi Ameghino, 1889 (Huayquerian) 25 19. Perathereutes pungens Ameghino, 1891c (Santacrucian) 27 20. Perathereutes pungens Ameghino, 1891c (Santacrucian) 28 21. Perathereutes pungens Ameghino, 1891c (Santacrucian) 28 22. Borhyaenidium musteloides Pascual & Bocchino, 1963 (Huayquerian) 32 23. Borhyaenidium musteloides Pascual & Bocchino, 1963 (Huayquerian) 34 24. Borhyaenidium musteloides Pascual & Bocchino, 1963 (Huayquerian) 35 25. Borhyaenidium musteloides Pascual & Bocchino, 1963 (Huayquerian) 36 26. Borhyaenidium riggsi sp. nov. (Montehermosan) 37 27. Borhyaenidium riggsi sp. nov. (Montehermosan) 38 28. Borhyaenidium riggsi sp. nov. (Montehermosan) 39 29. Borhyaenidium riggsi sp. nov. (Montehermosan) 40 30. Sipalocyon externa (Ameghino, 1902c) (Colhuehuapian) 41 31. Sipalocyon externa (Ameghino, 1902c) (Colhuehuapian) 42 32. Sipalocyon gracilis Ameghino, 1887 (Santacrucian) 45 33. Sipalocyon gracilis Ameghino, 1887 (Santacrucian) 45 34. Sipalocyon gracilis Ameghino, 1887 (Santacrucian) 46 35. Sipalocyon gracilis Ameghino, 1887 (Santacrucian) 46 36. Sipalocyon gracilis Ameghino, 1887 (Santacrucian) 47 37. Sipalocyon gracilis Ameghino, 1887 (Santacrucian) 48 38. Sipalocyon gracilis Ameghino, 1887 (Santacrucian) 52 39. Sipalocyon gracilis Ameghino, 1887 (Santacrucian) 54 40. Sipalocyon gracilis Ameghino, 1887 (Santacrucian) 55 41. Sipalocyon gracilis Ameghino, 1887 (Santacrucian) 56 42. Sipalocyon gracilis Ameghino, 1887 (Santacrucian) 58 43. Sipalocyon gracilis Ameghino, 1887 (Santacrucian) 59 44. Sipalocyon obusta (Ameghino, 1891c) (Santacrucian) 60 45. Sipalocyon obusta (Ameghino, 1891c) (Santacrucian) 61 46. Sipalocyon obusta (Ameghino, 1891c) (Santacrucian) 61 47. Notocynus hermosicus Mercerat, 1891b (Montehermosan) 64 vi vii 48. Notoq/nus hermosicus Mercerat, 1891b (Montehermosan) 65 49. Notogale mitis (Ameghino, 1897) (Deseadan) 66 50. Notogale mitis (Ameghino, 1897) (Deseadan) 67 51. Notogale mitis (Ameghino, 1897) (Deseadan) 69 52. Notogale mitis (Ameghino, 1897) (Deseadan) 70 53. Notogale mitis (Ameghino, 1897) (Deseadan) 71 54. Notogale mitis (Ameghino, 1897) (Deseadan) 71 55. Notogale mitis (Ameghino, 1897) (Deseadan) 72 56. Cladosictis centralis Ameghino, 1902c (Colhuehuapian) 74 57. Cladosictis centralis Ameghino, 1902c (Colhuehuapian) 75 58. Cladosictis centralis Ameghino, 1902c (Colhuehuapian) 77 59. Cladosictis centralis Ameghino, 1902c (Colhuehuapian) 78 60. Cladosictis patagonica Ameghino, 1887 (Santacrucian) 80 61. Cladosictis patagonica Ameghino, 1887 (Santacrucian) 81 62. Cladosictis patagonica Ameghino, 1887 (Santacrucian) 82 63. Cladosictis patagonica Ameghino, 1887 (Santacrucian) 83 64. Cladosictis patagonica Ameghino, 1887 (Santacrucian) 84 65. Cladosictis patagonica Ameghino, 1887 (Santacrucian) 86 66. Cladosictis patagonica Ameghino, 1887 (Santacrucian) 90 67. Cladosictis patagonica Ameghino, 1887 (Santacrucian) 91 68. Chasicostylus castroi Reig, 1957 (Chasicoan) 96 69. Chasicostylus castroi Reig, 1957 (Chasicoan) 97 70. Chasicostylus castroi Reig, 1957 (Chasicoan) 98 71. Chasicostylus castroi Reig, 1957 (Chasicoan) 99 72. Anatherium herrerae sp. nov. (Colhuehuapian) 100 73. Anatherium herrerae sp. nov. (Colhuehuapian) 101 74. Anatherium defossus Ameghino, 1887 (Santacrucian) 104 75. Anatherium defossus Ameghino, 1887 (Santacrucian) 105 76. Anatherium defossus Ameghino, 1887 (Santacrucian) 106 77. Anatherium defossus Ameghino, 1887 (Santacrucian) 107 78. Anatherium defossus Ameghino, 1887 (Santacrucian) 107 79. Comparison of upper dentitions of various species of Hathlyacyninae showing relative size and proportions of teeth 109 80. Comparison of lower dentitions of various species of Hathlyacyninae showing relative size and proportions of teeth 110 81. Size distribution of various species of Hathlyacyninae as indicated by length of MM 112 82. Size distribution of various species of Hathlyacyninae as indicated by relationship of length and width of M, 114 83. Dendrogram showing probable phylogenetic relationships of the genera and species of Hathlyacyninae 115 LIST OF TABLES 1. Measurements of cheek teeth of Patene simpsoni and P. coluapiensis 12 2. Measurements of cheek teeth of Pseudonotictis pusillus and Notictis ortizi 21 3. Measurements of mandibular rami of Pseudonotictis pusillus and Notictis ortizi . 21 4. Measurements of lower cheek teeth and mandibular rami of Perathereutes pungens 29 5. Measurements of cheek teeth of Borhyaenidium musteloides, B. riggsi sp. nov, and Notocynus hermosicus 33 6. Measurements of mandibular rami of Borhyaenidium musteloides, B. riggsi sp. nov., and Notocynus hermosicus 34 7. Measurements of upper dentition of Sipalocyon externa 43 8. Measurements of mandibular rami of Sipalocyon gracilis 49 9. Statistics for some dimensions of mandibular rami of Sipalocyon gracilis 49 10. Measurements of upper cheek teeth of Sipalocyon gracilis 50 11. Statistics for some upper cheek tooth dimensions of Sipalocyon gracilis 51 12. Measurements of lower cheek teeth of Sipalocyon gracilis 53 13. Statistics for some lower cheek tooth dimensions of Sipalocyon gracilis 54 14. Measurements of cheek teeth and mandibular rami of Sipalocyon obusta 62 15. Measurements of cheek teeth of Notogale mitis 68 16. Measurements of lower cheek teeth of Cladosictis centralis 76 17. Measurements of mandibular rami of Cladosictis patagonica 81 18. Statistics for some dimensions of mandibular rami of Cladosictis patagonica .... 82 19. Measurements of upper cheek teeth of Cladosictis patagonica 88 20. Statistics for some upper cheek tooth dimensions of Cladosictis patagonica 89 21 . Measurements of lower cheek teeth of Cladosictis patagonica 92 22. Statistics for some lower cheek tooth dimensions of Cladosictis patagonica 94 23. Measurements of cheek teeth of Chasicostylus castroi 97 24. Measurements of lower cheek teeth of Anatherium defossus and A. herrerae sp. nov 102 25. Measurements of mandibular rami of Anatherium defossus and A. herrerae sp. nov 103 26. Summary of some diagnostic characters for genera of Hathlyacyninae Ill ABSTRACT Members of the extinct "dog-like" marsupial subfamily HATHLYACYNINAE (Borhyaenidae, Borhyaenoidea) are known from beds of Late Paleocene (Rio- chican) through Pliocene (Montehermosan) age in Argentina and beds of early Oligocene (Deseadan) age in Bolivia. Twelve genera and 18 species are recog- nized: Patene simpsoni Paula Couto, 1952; Patene coluapiensis Simpson, 1935; Pro- cladosictis anomala Ameghino, 1902; Pseudonotictis pusillus (Ameghino, 1891) gen. nov.; Notictis ortizi Ameghino, 1889; Perathereutes pungens Ameghino, 1891; Bor- hyaenidium musteloides Pascual & Bocchino, 1963; Borhyaenidium riggsi sp. nov.; Sipalocyon externa (Ameghino, 1902); Sipalocyon gracilis Ameghino, 1887; Sipalo- cyon obusta (Ameghino, 1891); Notocynus hermosicus Mercerat, 1891; Notogale mitis (Ameghino, 1897); Cladosictis centralis Ameghino, 1902; Cladosictis patagonica Ame- ghino, 1887; Chasicostylus castroi Reig, 1957; Anatherium herrerae sp. nov.; and Anatherium defossus Ameghino, 1887. In addition, Procladosictis erecta Ame- ghino, 1902; Pseudocladosictis determinabile Ameghino, 1902; and Notogale tenuis (Ameghino, 1897) are regarded as nomina vana. The species and genera are distinguished largely on the basis of absolute and relative size differences in the dentition. Such characters as presence or absence of a metaconid; relative size of protocone, talonid, and stylar shelf; spacing differences between C, P„ and P2; and orientation of P, in the jaw relative to other cheek teeth also proved useful in distinguishing these taxa. With regard to dental structure and incisor number, hathlyacynes are the most generalized of known borhyaenids, and early members (i.e., Patene) closely ap- proximate the expected condition of their presumed didelphoid ancestors. The general evolutionary trend within the Hathlyacyninae involved increase in car- nassial specializations resulting in loss of metaconid and reduction in size of protocone, stylar shelf, and talonid. The group has been extremely conservative during its evolutionary history, and structurally the taxa are monotonously alike. Nevertheless, if sheer numbers of individuals and taxa are a gauge, then the Hathlyacyninae was the most successful of the borhyaenid subfamilies. INTRODUCTION This paper presents a detailed systematic revision of an extinct subfamily of South American "dog-like" marsupials, the Hathlyacyninae. The taxonomic history of the Hathlyacyninae is reviewed, the possible phylogenetic relation- ships of the included taxa are discussed, and the group's taxonomy at the generic and specific levels is stabilized. This study represents an attempt to bring to- gether in one place an updated and expanded treatment of these animals. Sys- 2 FIELDIANA: GEOLOGY tematic revisions of other extinct carnivorous marsupial groups in South America are presented elsewhere [Thylacosmilidae (Marshall, 1976b); Borhyaenidae in general and Borhyaeninae in particular (Marshall, 1978); Prothylacyninae (Mar- shall, 1979)]. During the course of this study I was able to examine, firsthand, all pertinent materials, including type and referred specimens. This work includes discussion and description of some new materials, including two new species, but is es- sentially based on a reappraisal of previously known specimens and literature. The fossil localities mentioned below (figs. 1, 2) are shown on maps and are discussed in detail in Marshall (1976c) and in Marshall, Hoffstetter, & Pascual (in press). The chronology and usage of South American Land Mammal Ages follows Marshall, Hoffstetter, & Pascual (in press), which is based largely on the radioisotope time scale in Marshall, Pascual, et al. (1977) and Marshall, Butler, et al. (1979). All measurements are in millimeters (mm) unless indicated oth- erwise. The serial designation for cheek tooth number employed in this study is based on the dental formula P{^, M{^. ABBREVIATIONS Abbreviations used in text, figure captions, and tables of measurements are: C, canine; ca., approximate measurement; CV, coefficient of variation; I, incisor; L, length; M, molar; N, number; OR, observed range of sample; P, premolar; s, standard deviation of sample; x, mean; W, width. The following abbreviations are used for specimens from institutional collec- tions: AC, Amherst College, Amherst, Mass.; AMNH, American Museum of Natural History, New York; BM(NH), British Museum (Natural History), Lon- don; CMNH, Carnegie Museum of Natural History, Pittsburgh; DGM, Divasao de Geologia e Mineralogia do Departamento Nacional da Producao Mineral, Rio de Janeiro, Brazil; FMNH, Field Museum of Natural History, Chicago; MACN, Museo Argentino de Ciencias Naturales "Bernardino Rivadavia," Buenos Aires, Argentina; MLP, Museo de La Plata, La Plata, Argentina; MMP, Museo Municipal de Ciencias Naturales de Mar del Plata "Lorenzo Scaglia," Mar del Plata, Ar- gentina; MNHN, Museum National d'Histoire Naturelle, Paris, France; MNRJ, Museu Nacional e Universidade Federal do Rio de Janeiro, Brazil; PU, Princeton University, Princeton, N.J.; UCMP, University of California Museum of Paleon- tology, Berkeley, Calif.; USNM, United States National Museum, Washington, D.C ACKNOWLEDGMENTS I am indebted to the following individuals for allowing me to study specimens at their respective institutions: D. Baird (PU), R. H. Tedford (AMNH), R. Pascual (MLP), J. Bonaparte (MACN), F. L. de Souza Cunha (MNRJ), L. Price (DGM), A. J. Sutcliff [BM(NH)], R. Hoffstetter (MNHN), M. Coombs (AC), and M. Dawson (CMNH). For many helpful comments on various phases of this study and/or for reading the manuscript, I thank M. C. McKenna, R. Pascual, R. H. Tedford, and W D. Turnbull. The figures of specimens (except figs. 49, 51-55) were drawn by Mar- lene Hill Werner from original specimens and/or epoxy casts. The latter were Fie. 1. Map of South America showing some vertebrate fossil localities (circles) discussed in text. OCEANO PACI FICO 0 50 100 150 200 k Fig. 2. Map of southern tip of South America showing vertebrate fossil localities (circles) discussed in text. MARSHALL: REVIEW OF THE HATHLYACYNINAE 5 made by Mrs. Barbara Waters at UCMR The stereo photographs were made by Ron Testa from epoxy casts. Initial stages of this study were facilitated by support from UCMP for travel and graduate study and by grants no. 1329, 1698, 1943 from the National Geo- graphic Society, Washington, D.C.; its completion was made possible by National Science Foundation grant DEB-7901976. SYSTEMATICS Superfamily BORHYAENOIDEA (Ameghino, 1894, p. 371) Simpson, 1930, p. 9. Diagnosis. — Dental formula IJH, Q, P£i M$; extinct South American "dog- and cat-like" marsupials of small to large size; lack palatal vacuities; transverse canal either rudimentary or absent; strong sagittal and nuchal crests; lunar small and in contact with large magnum; lacrimal bone extends onto rostrum and usually has large tuberosity developed above lacrimal canal which opens within orbit. Known range. — Riochican through Montehermosan. Family BORHYAENIDAE Ameghino, 1894, p. 371 Diagnosis. — Dental formula l£f, C\, P], M}; small to large size; skull dolicho- cephalic to brachycephalic, rostrum robust and well developed; upper and lower C usually large, laniary, and with closed roots in adults; mandibular symphysis typically shallow (may be fused or unfused in adult) and without flange; man- dibular ramus of subequal depth and breadth below molar series and without distinct labial bend posteriorly along ventral edge as in thylacosmilids; masseteric fossa usually shallow; premolars double rooted; molars increase gradually or rapidly in size from M{ to M$; protocone large to very reduced; paracone often reduced; paracone and metacone approximated on M1'3; stylar shelf reduced; talonids large or reduced, and often imperfectly or not basined; metaconids often absent, if present always smaller than paraconids; nasals large and expanded posteriorly; distinct nasal-lacrimal contact; no postorbital bar; basicranial and basifacial planes parallel; basisphenoid and basioccipital processes increase in width posteriorly, neither has a distinct medial keel and both are relatively flat transversely; at suture they form a fairly prominent transverse ridge; pars petrosa of periotic lacks a tympanic process; large hypoglossal, postsquamosal and post- glenoid foramina present. Known range. — Riochican through Montehermosan. Subfamily HATHLYACYNINAE (Ameghino, 1894, p. 382) Kirsch, 1977, p. 112 [Including Acyonidae,1 Ameghino, 1889, p. 894; 1891a, p. 147n; Hathlyacynidae Ameghino, 1894, p. 382; Amphiproviverridae Ameghino, 1894, p. 333n (formally proposed on p. 389); Cladictidae Winge, 1923, p. 77; Cladosictidae Ameghino, 1935, p. 131; Cladosictinae Cabrera, 1927, p. 273) 'The family-group name Acyonidae was formally proposed by Ameghino (1891a, p. 147n) to accommodate Acyon and Sipalocyon. The name first appears in a list of families in Ameghino (1889, p. 894). Trouessart (1898, p. 1215) later recognized this taxon as a 6 FIELDIANA: GEOLOGY Diagnosis. — It C{, Pi M| (when known); small to medium-sized borhyaenids; mandibular rami long, shallow, and generally gracile; symphysis ligamentous and rami never fused in adult; symphysis typically extends posteriorly to point below P3; IM subequal in size; C weakly or moderately developed; P{ much smaller than P| or P^; P2. and Pj either subequal in size or latter is larger; P13 usually have small but distinct posterobasal cusps; molars increase in size from Mj to M4; metaconid present only in earliest forms (Patene); M13 have large, basined talonid (M4 with reduced talonid that may or may not be basined); small but distinct anterobasal cingulum on M2^, weak or absent on M,; Iw of subequal size; M'^with well-developed protocone; paracone well developed and distinct on M1'3, but smaller and fused basally with larger metacone; weak but distinct parastyle on M1"3, increasing in size from M1 to M3; distinct paracingulum connects parastyle with protocone; metacrista moderately well developed; M3 usually with distinct ectoflex, small ectoflex sometimes present on M2; skull dolichocephalic; large ossified auditory bulla (alisphenoid forms anterior two-thirds, pars mas- toidea of periotic forms posterior one-third — Patterson, 1965); ectotympanic has large incisura tympanica dorsally; epitympanic recess rather large and circular; paroccipital process heavy, squat, and blunt; pars petrosa of periotic rather sharply pointed anteriorly; foramen ovale, foramen lacerum posterium, and posterior carotid foramen all large; foramen lacerum medium absent; terminal phalanges uncleft, laterally compressed, and pointed. Known range. — Riochican through Monte hermosan. Patene Simpson, 1935 Patene Simpson, 1935, p. 3. Ischyrodidelphis Paula Couto, 1952b, p. 9. Type of Patene. — Patene coluapiensis Simpson, 1935, p. 3. Type of Ischyrodidelphis. — Ischyrodidelphis castellanosi Paula Couto, 1952b, p. 11. Distribution. — Riochican of Brazil and Casamayoran of Argentina. Diagnosis. — Borhyaenids of small size; M1^ with large, well-developed pro- tocone and basined talon; paracone slightly smaller than and well separated from metacone on M1 3 (both cusps are more external on M1 and medial on M3); metastylar spur weakly developed; M1 3 with distinct but vestigial protoconules subfamily, Acyoninae, and included it with the Borhyaenidae. The last instance in which the family-group name Acyonidae is employed as a valid name appears in Trouessart (1904, p. 176), who recognized it as distinct from the family-group name Borhyaenidae, and in Palmer (1904, p. 877), who included it as a synonym of the family-group name Borhyaenidae. Palmer (1904, p. 877n) noted that "Acyonidae has priority of five years merely by publication in a nominal list, but as Borhyaenidae has come into more general use it is here adopted provisionally." The family-group name Acyonidae has remained unused as a senior synonym in all zoological literature for the last 70 years. During this time, the family-group name Bor- hyaenidae has always been used, and Acyonidae has been ignored by everyone working on members of this family. In addition, Acyon Ameghino, 1887, is a junior synonym of Cladosictis Ameghino, 1887 (Cabrera, 1927, p. 288; this study). A proposal has been sub- mitted to the Commission requesting use of its plenary powers to suppress the family- group name Acyonidae Ameghino, 1889, for the purposes of the Law of Priority, but not for those of the Law of Homonymy, and to place this family-group name on the Official Index of Rejected and Invalid Family-Group Names in Zoology (Marshall, Clemens, et al., 1977, 1978). MARSHALL: REVIEW OF THE HATHLYACYNINAE 7 and metaconules; shallow but distinct ectoflex present on M5; M4with strong parastylar spur, paracone median, metacone represented by basal cuspule; stylar shelf well developed on M1 3; parastyle small but distinct on M1 2, weakly de- veloped on MJ; P} proodont with distinct posterobasal cuspule; trigonids prom- inent and much higher than talonids; metaconid well developed on M2^, smaller on M,; talonids well developed and basined on MM and narrower than trigonids; M,., with well-developed anterobasal cingular shelf; no evidence of diastems between C, P„ and/or P2. Patene simpsoni Paula Couto, 1952a. Figures 3-8; Table 1. Patent simpsoni Paula Couto, 1952a, p. 23, fig. 8A, B; 1961, p. 329, figs. 8-10; 1970, p. 33. Ischyrodidelphis castellanosi Paula Couto, 1952b, p. 11; 1962, p. 140, fig. 1. Type of Patene simpsoni. — MNRJ 1331-V , a right maxillary with P3-M4 (figured by Paula Couto, 1961, figs. 8-9). Type of Ischyrodidelphis castellanosi. — MNRJ 1351-V, a partial right mandibular ramus with roots of P2, P3-M2, and trigonid of M3 present, and alveoli of M, talonid and M< (figured by Paula Couto, 1962, fig. 1A-B). Fig. 3. Patene simpsoni Paula Couto, 1952a (Riochican). MNRJ 1331-V (type), a right maxillary with P3-M4 complete: a, labial; b, occlusal; c, lingual views. Scale = 10 mm. i* ;&**• Fig. 4. Patene simpsoni Paula Couto, 1952a (Riochican). Stereopairs of MNRJ 1331-V (type), a right maxil- lary with P3-M4 complete: a, labial; b, occlusal; c, lingual views. Scale = 10 mm. MARSHALL: REVIEW OF THE HATHLYACYNINAE Fie. 5. Putene simpsoni Paula Couto, 1952a (Riochican). MNRJ 1351-V (type of Ischyrodi- delphis castellanosi Paula Couto, 1952b), a partial right mandibular ramus with roots of P2, P3-M2 and trigonid of M3 present, and alveoli of M3 talonid and M,: a, labial; b, occlusal; c, lingual views. Scale = 30 mm. Hypodigm. — The two types and DGM 324-M, a fragment of a right mandibular ramus with alveoli of P: and roots of P3, anterior root of M„ talonid and anterior root of M2, M3 virtually complete, and roots of M4 (the M3 is figured by Paula Couto, 1952a, fig. 8A, B; 1961, fig. 10); DGM 331-M, part of a left maxilla with alveoli of M1'3 and incomplete alveoli of P3 and M4; DGM 654-M, an isolated right Mv DGM 655-M, an isolated right M3; DGM 656-M, an isolated right M,; DGM 797-M, an isolated right M3; DGM 798-M, an isolated left M,; MNRJ 1332-V, an isolated right M4; MNRJ 1333-V (now AMNH 49804), an isolated left M1; MNRJ 1334-V (now AMNH 49805), an isolated right M2; MNRJ 1335-V, an incomplete isolated M1 or M2; MNRJ 1336-V, an isolated left lower trigonid; MNRJ 1337-V, an isolated left lower M4 trigonid; MNRJ 1338-V , an isolated left lower trigonid; MNRJ 1339-V, an isolated left lower trigonid; MNRJ 1340-V, an isolated left lower trigonid; MNRJ 1341-V, an isolated left M3 trigonid; MNRJ 1342-V, an isolated left lower trigonid; MNRJ 1343-V, an isolated left lower trigonid; MNRJ 1352-V, an isolated right M^ MNRJ 1353-V, an isolated right M: (figured by Paula Couto, 1962, fig. 1C-E); MNRJ 1354-V, an isolated left M^; MNRJ 1425-V, an isolated right Pv- MNRJ 1428-V, an isolated right lower molar trigonid; (DGM 324-M, MNRJ 1332-V, and MNRJ 1335-V are paratypes of P. simpsoni; and MNRJ 1352-V, 1353-V, 1354-V, and 1425-V are paratypes of /. castellanosi). Horizon and locality. — All specimens were collected from fissure deposits in the Itaboraf Formation, Sao Jose de Itaboraf, Brazil. Age. — Riochican . Fig. 6. Patene simpsoni Paula Couto, 1952a (Riochican). Stereopairs of MNRJ 1351-V (type of Ischyrodidelphis castellanosi Paula Couto, 1952b), a partial right mandibular ramus with roots of P2, P3-M2 and trigonid of M3 present, and alveoli of M3 talonid and M4: a, labial; b, occlusal; c, lingual views. Scale = 30 mm. 10 MARSHALL: REVIEW OF THE HATHLYACYNINAE 11 Fie. 7. Patent simpsoni Paula Couto, 1952a (Riochican). DGM 324-M, a fragment of a right mandibular ramus with alveoli of P2 and roots of Pj, anterior root of Mw talonid and anterior root of M2, M3 virtually complete, and roots of M,: a, labial; b, occlusal; c, lingual views. Scale = 20 mm. Diagnosis. — Differs from Patene coluapiensis in being about one-fourth to one- third smaller in linear tooth dimensions and in having relatively larger proto- cones and wider stylar shelf on M1 "\ a larger metacone on M\ and a slightly shallower ectoflex on M\ Comments. — Paula Couto (1952b, p. 11) erected Ischyrodidelphis castellanosi on a partial lower dentition and several isolated cheek teeth. He assigned this species to the subfamily Didelphinae and concluded (1952b, p. 9) that "its closest affinities seem to be with the largest Recent species of the genus Didelphis." I have compared all specimens assigned by Paula Couto to /. castellanosi with the lower dentitions of Patene simpsoni and found them to represent the same species. I therefore formally recognize /. castellanosi as a junior synonym of P. simpsoni. The recognition that these species are synonymous readily explains why Paula Couto did not assign upper cheek teeth to /. castellanosi; they were all assigned to P. simpsoni. In the lower dentition, he assigned complete cheek teeth from M2 forward to /. castellanosi, whereas he assigned complete M3s and M4s to P. simpsoni. Patene simpsoni is the most generalized of known borhyaenoids. Indeed, if more specialized borhyaenoids like cf. Nemolestes sp. were not contemporaneous with it (see Marshall, 1978, p. 27), P. simpsoni would be an ideal prototype for the superfamily. Patene simpsoni is, nonetheless, structurally the most generalized and temporally the oldest of known Hathlyacyninae and may represent the basal stock for the subfamily. £ l-> vo in in en CD CD o On i— O O i-i • ON O tv CN CD CD CD CD • CM CD r^ vOOOOOON t^ in m so m in V l-l a 2 *? in o o 00 00 ON ON s 1 E << X >>>> oc M a. en w H M w w X u R 5 & 5 7, MNRJ 1331- MNRJ 1332- MNRJ 1333- MNRJ 1334- « S3 e .§ 5. s _s 3 a CM X 2 2 < s M w w X o M B E D Qi Qi C 22222 > >> *•* invoooSSS •sCNiinininoNS22S S_2 22 22!,<i<f •KQQDDD222 12 MARSHALL: REVIEW OF THE HATHLYACYNINAE 13 Fig. 8. Patene simpsoni Paula Couto, 1952a (Riochican). Stereopairs of DGM 324-M, a fragment of a right mandibular ramus with alveoli of P2 and roots of P3, anterior root of M, , talonid and anterior root of M2, M3 virtually complete, and roots of M,: a, labial; b, occlusal; c, lingual views. Scale for labial and lingual views = 30 mm. Patene coluapiensis Simpson, 1935. Figures 9, 10; Table 1. Patene coluapiensis Simpson, 1935, p. 3; 1948, pp. 44-45, fig. 5. Type. — AMNH 28448, a right maxillary with MM (figured by Simpson, 1948, fig. 5). Hypodigm. — Type only Horizon and locality. — From Casamayor beds south of Lago Colhu£-Huapi, Chubut Province, Argentina; it is from the lowest known fossil horizon of this age. Age. — Casamayoran. Diagnosis. — Differs from P. simpsoni in being about one-fourth to one-third larger in linear tooth dimensions and in having relatively smaller protocones and a slightly narrower stylar shelf on M'\ a more reduced metacone on M\ and a slightly deeper ectoflex on M\ Comments. — Patene coluapiensis is very similar to the Riochican P. simpsoni, and except for the features noted in the diagnosis above, these species are insepar- able. Patene coluapiensis is known only from the Casamayoran of Argentina, and P. simpsoni, from the Riochican of Brazil. Considering their structural similarity and their occurrence in successive aged beds, they may be regarded as a single evolutionary lineage. I therefore recognize P. coluapiensis as a slightly larger, more specialized descendant of P. simpsoni. 14 FIELDIANA: GEOLOGY Fig. 9. Patene coluapiensis Simpson, 1935 (Casamayoran). AMNH 28448 (type), a right maxillary with M1"4: a, labial; b, occlusal; c, lingual views. Scale = 20 mm. Simpson (1948, p. 45) noted that an isolated broken upper molar (AMNH 28532) from the Rio Chico beds at Canadon Hondo was similar to P. coluapiensis and accordingly he referred it to Patene sp. This assignment is, however, very tentative, and it could just as convincingly be argued that this specimen was referable to other Casamayoran genera such as Nemolestes or Argyrolestes, both poorly known members of the subfamily Borhyaeninae (see Marshall, 1978, pp. 26-27). My preference is to regard AMNH 28532 as Borhyaenidae, genus and species indeterminate. Fig. 10. Fatene coluapiensis Simpson, 1935 (Casamayoran). Stereopairs of AMNH 28448 (type), a right maxillary with M ' a, labial; b, occlusal; c, ungual views. Scale = 20 mm. 15 16 FIELDIANA: GEOLOGY Procladosictis Ameghino, 1902b Procladosictis Ameghino, 1902b, p. 46; 1906, p. 354; Schlosser, 1923, p. 440; Scott, 1937, p. 704. Type. — Procladosictis anomala Ameghino, 1902b, p. 46. Distribution. — Mustersan beds, Patagonia, southern Argentina. Diagnosis. — As for type and only known species. Procladosictis anomala Ameghino, 1902b. Figures 11, 12. Procladosictis anomala Ameghino, 1902a, p. 320 (nomen nudum); 1902b, p. 46; 1906, p. 354, fig. 191; Simpson, 1948, p. 45, pi. 2, fig. 7. Type. — MACN 10327, a fragment of a right maxillary with P3-M3 complete (figured by Ameghino, 1906, fig. 191; Simpson, 1948, fig. 7). Fig. 11. Procladosictis anomala Ameghino, 1902b (Mustersan). MACN 10327 (type), a fragment of a right maxillary with P3-M3 complete: a, labial; b, occlusal; c, lingual views. Scale = 20 mm. MARSHALL: REVIEW OF THE HATHLYACYNINAE 17 '±m+ Fig. 12. Procladosictis anomala Ameghino, 1902b (Mustersan). Stereopairs of MACN 10327 (type), a fragment of a right maxillary with P^-M3 complete: a, labial; b, occlusal; c, lingual views. Scale = 30 mm. Hypodigm. — Type only. Horizon and locality. — "Couches a Astraponotus," Chubut Province, Patagonia, southern Argentina. No other data. Age. — Mustersan. Diagnosis. — Small borhyaenid; crown of P3 higher than M1, and with a distinct posterobasal heel; protocone distinct but low on M1'3, and becoming smaller from M1 to M3 (very weak on M3); talon cuspate; paracone and metacone con- fluent basally; metacone slightly larger than paracone on M1, paracone smaller on M2 and very reduced on M3; metacone in turn becomes progressively larger from M1 to M3; paracone and metacone median on M\ progressively more in- ternal on M23; outer margin nearly straight on M'"2, and bilobed with deep ectoflex on M3; twin styles, anterior larger and external to paracone and metacone on M:, small on M1, absent on M3; distinct paracingulum on M2, weak on M1 and M3 (modified from Simpson, 1948, p. 45). Measurements.— L P3 = 9.0 mm, W P3 = 4.5 mm; LM' = 9.0 mm, W M1 = 6.5 mm; L M2 = 9.5 mm, W M2 = 8.0 mm; L M3 = 7.5 mm, W M3 = 10.0 mm. Comments. — The P3 was restored onto the anterior edge of the M1 with wax. The oblique position that this tooth now occupies, as noted by Simpson (1948, 18 FIELDIANA: GEOLOGY p. 45), is thus a factor of restoration and is not a diagnostic character of the species or genus. The anterolabial corner of the M2 is broken off and with it parts of the two stylar cusps. The P3-M2 of Procladosictis anomala are generalized in structure and resemble not only species of Patene from the Riochican and Casamayoran but also later taxa of similar size such as Sipalocyon and Cladosictis from the Colhuehuapian and Santacrucian. The M3 of P. anomala has a very deep ectoflex, a feature not found in other Borhyaenidae. This feature makes P. anomala very distinct, and represents a specialization that excludes it as an ancestor for any known later taxa. Based on present evidence, P. anomala is regarded as a dead-end offshoot of an early Tertiary hathlyacynine radiation. It apparently represents a slightly larger specialized descendant of a population of Patene coluapiensis. Simpson (1948, p. 45) recorded that a lower isolated left molar (AMNH 29433), possibly an M3 from Mustersan beds at Cerro del Humo, may belong to Procla- dosictis. The tooth measures L = 9.8 mm, W of trigonid = 4.9 mm, W talonid = 4.9 mm. The tooth has a minute metaconid and a narrow-basined talonid with hypoconulid poorly distinguished from entoconid. Procladosictis erecta Ameghino, 1902b Procladosictis erecta Ameghino, 1902a, p. 316 (nomen nudum); 1902b, p. 47. "Procladosictis" erecta Simpson, 1948, p. 45 (as nomen vanum). Type. — MACN 10328, a lower premolar. Hypodigm. — Type only. Horizon and locality. — "Partie superieure des couches a Notostylops," north of Lago Colhue-Huapi, Chubut Province, Argentina (for discussion on this locality see Simpson, 1967, p. 64). Age . —Casamayoran . Diagnosis. — Indeterminate. Comments. — "There is no real reason for referring this [species] to Procladosictis, a genus based on upper teeth from a different horizon. Nor does there seem to be any way in which this supposed species can now be defined or distinguished. The type is a lower premolar, 9.3 mm long and 4.0 mm wide, with central cusps and sloping, noncuspidate heel. It may be kept on record, since a neotype might show the species to be distinctive, but its present value and significance are nil" (Simpson, 1948, pp. 45-46). Pseudocladosictis Ameghino, 1902b Pseudocladosictis Ameghino, 1902b, p. 47. Type. — Pseudocladosictis determinable Ameghino, 1902b, p. 47. Distribution. — Casamayor beds, Patagonia. Diagnosis. — As for type and only known species. Pseudocladosictis determinabile Ameghino, 1902b Procladosictis determinabile Ameghino, 1902a, p. 316 (nomen nudum). Pseudocladosictis determinabile Ameghino, 1902b, p. 47; Simpson, 1948, p. 46 (as nomen vanum). Type. — MACN 10325, an isolated premolar, considered a lower by Ameghino, but an upper by Simpson. MARSHALL: REVIEW OF THE HATHLYACYNINAE 19 Hypodigm. — Type only. Horizon and locality. — "Couches a Notostylops" (upper part, fide Ameghino) from Great Barranca south of Lago Colhu£-Huapf, Chubut Province, Argentina. Age. — Casamayoran. Diagnosis . — Indetermina te . Comments. — "The type measures 8.8 by 6.1 mm and has a central cusp and small, partly transverse, cuspidate heel. Like Procladosictis erecta, the species [determinabile] is essentially no more than preemptive, based on no good evidence but securing the species to its author in case later work shows it to be valid" (Simpson, 1948, p. 46). Pseudonotictis gen. nov. Etymology. — Pseudo- (Gr., false or lie), name given in reference to morphological similarity to Huayquerian genus Notictis. Type. — Pseudonotictis pusillus (Ameghino, 1891c, p. 315). Distribution. — Santa Cruz Formation, Santa Cruz Province, Argentina. Diagnosis. — As for type and only known species. Pseudonotictis pusillus (Ameghino, 1891c). Figures 13-16; Tables 2, 3. Sipalocyon pusillus Ameghino, 1891c, p. 315; 1894, p. 393; 1935, p. Ill, fig. 12 (caption only); Cabrera, 1927, p. 293. Hathliacynus kobyi Mercerat, 1891a, p. 53. Perathereutes hobyi Cabrera, 1927, p. 293, figs. 9, 10. Type of Pseudonotictis pusillus. — MACN 666, a fragment of a right mandibular ramus with alveoli of I,3, C, and P12; anterior root of P3; posterior half of Pv anterior half of M,; posterior root of M, and anterior alveolus of M2. Type of Hathliacynus kobyi. — MLP 11-26, a left mandibular ramus with root of C, P, and P3-M4 complete, anterior half and posterior root of P2 present; fragment of a right mandibular ramus with P13 complete; fragment of a right maxillary with alveoli of P3-M', and a complete M2; and various cranial fragments; a right humerus (figured by Cabrera, 1927, fig. 10); right radius; part of a right ulna; and two metapodials; all of a single associated individual. [Cabrera (1927, fig. 9) figured as Perathereutes kobyi an edentulous right maxillary and a left mandib- ular ramus in which the M3 was missing the posterior half of the trigonid and all of the talonid. These elements are part of the type of Hathliacynus kobyi (MLP 11-26). In 1975, I found several isolated teeth and tooth fragments associated with these elements. An isolated right M2 fit perfectly into the alveolar coun- terpart of the maxillary fragment, and the missing portion of the lower left M3 was also present. Figures 13-16 presented here show these elements with the teeth restored.] Hypodigm. — The two types and MLP 11-40, a fragment of a right mandibular ramus with M,. Horizon and locality. — All specimens are from the Santa Cruz Formation, Santa Cruz Province, Argentina. MACN 666 is from Monte Observaci6n and was collected by Carlos Ameghino in 1890-91. MLP 11-26 and 11-40 are from Monte Leon (no other data). Age. — Santacrucian. 20 FIELDIANA: GEOLOGY Fig. 13. Pseudonotictis pusillus (Ameghino, 1891c) (Santacrucian). MLP 11-26 (type of Hathliacynus kobyi Mercerat, 1891a), a left mandibular ramus with root of C, P, and P3-M4 complete, and anterior half and posterior root of P2 present: a, labial; b, occlusal; c, lingual views. Scale = 30 mm. Diagnosis. — Smallest known species of Santacrucian borhyaenid; C weakly developed; large, elongated mental foramen below P12 contact and below M,; Pt very reduced compared with P2 3; MM talonids well developed and basined, but proportionately smaller than in contemporary species of Sipalocyon; M4 tal- onid very reduced but basined; P,.3 aligned in same anteroposterior axis; P] separated from C and P2 by small diastems; posterobasal heel not distinct on Pi_3; M2 with distinct protocone, talon not basined; paracone well developed and distinct, but smaller and fused basally with large metacone; a small but distinct ectocingulum and parastyle present on M2, a very shallow ectoflex occurs labial to metacone; differs from Notictis ortizi in having proportionately smaller tal- onids, in an absolutely longer premolar region and a larger Px. Comments. — Sipalocyon pusillus was erected by Ameghino in 1891 (1891c, p. 315). That same year Mercerat (1891a, p. 53) described Hathliacynus kobyi. Three years later Ameghino (1894, p. 393) formally recognized H. kobyi as a junior synonym of S. pusillus and placed them in the Sparassodonta, family Amphi- proviverridae. Cabrera (1927, p. 393, figs. 9, 10), however, recognized H. kobyi as a species of Perathereutes and regarded S. pusillus as valid. In the present study I recognize the type species of Sipalocyon, S. gracilis, as valid, and the type species of Hathliacynus, H. lustratus, as a junior synonym of Cladosictis patagonica. Perathereutes pungens is also here recognized as valid. T 2 *- J- t>s © !s ! fl vO * in ro rvi «N 06 •-: S ir, 3 ^- e f ^ 3 s <5 en vo in X 3 § i - 3 — — — — ° S ° S\2's - E •G I a. I/} 06 ts! 3 Z «-" ... Z ?Ucu ay I E I 55" 5T ^= a. B 8? I § 2 S S .H 2 g == U a, a. • a 0 s §.2ss ?S 21 22 FIELDIANA: GEOLOGY Fig. 14. Pseudonotictis pusillus (Ameghino, 1891c) (Santacrucian). Stereopairs of MLP 11- 26 (type of Hathliacynus kobyi Mercerat, 1891a), a left mandibular ramus with root of C, P, and P3-M4 complete, and anterior half and posterior root of P2 present: a, labial; b, occlusal; c, lingual views. Scale = 30 mm. If the species pusillus is to be included in any presently recognized genus following arrangements of previous workers, it would have to be placed either in Cladosictis, from which it is certainly distinct and does not belong, or in Sipalocyon as originally proposed by Ameghino. There are sufficient differences, however, between pusillus and species of Sipalocyon to warrant generic separa- tion. For this reason I have proposed the new generic name Pseudonotictis. Pseudonotictis pusillus differs from species of Sipalocyon in being smaller in size, in having a proportionately smaller Pu in lacking a distinct (or less prominent) posterobasal heel on P,.3, in having a relatively and absolutely smaller protocone on upper molars, and in having a more weakly developed canine. Pseudonotictis pusillus is similar in size to the Riochican species Patene simpsoni. These species differ in P. pusillus having a more specialized dentition in the MARSHALL: REVIEW OF THE HATHLYACYNINAE 23 Fie. 15. Pseudonotictis pusillus (Ameghino, 1891c) (Santacrucian). MLP 11-26 (type of Hathlmynus kobyi Mercerat, 1891a), a frag- ment of a right maxillary with alveoli of P- M ; . and a complete M:: a, labial; b, occlusal; c, lingual views. Scale = 5 mm. direction of carnassialization than does P. simpsoni. The primary changes in- volved in such a lineage include reduction in size of the protocone, stylar shelf, talonid, and metacone (latter becoming fused basally with paracone); virtual loss of posterobasal cuspule on Pv loss of metaconid; elongation and tricuspid struc- ture of M,; and an increase in size and importance of postvallum-prevallid shear. Considering the time involved, Riochican to Santacrucian, these changes are minor, and I thus choose to regard P. simpsoni as the Riochican structural ancestor of P. pusillus. Pseudonotictis pusillus is in turn very similar to Notictis ortizi from the Huay- querian. Both species are of similar size, both have virtually identical molar morphology, both have weakly developed canines, and the size, shape, and location of the mental foramina are similar. The premolars in N. ortizi are ab- solutely and relatively smaller than in P. pusillus, and the premolar region of the jaw is more shortened and the teeth crowded, such that the P, is set obliquely in the jaw. These differences are related to shortening of the premolar region of the jaw and are thus the expression of a single basic change. Even taken as a whole, the differences between these species are not great and represent only slight modification or specialization of features seen in P. pusillus. An ancestral- descendant relationship for these taxa seems probable. 24 FIELDIANA: GEOLOGY tT2r*Jii Fig. 16. Pseudonotictis pusillus (Ameghino, 1891c) (Santacru- cian). Stereopairs of MLP 11-26 (type of Hathliacynus kobyi Mer- cerat, 1891a), a fragment of a right maxillary with alveoli of P3- M1, and a complete M2: a, labial; b, occlusal; c, lingual views. Scale = 10 mm. Notictis Ameghino, 1889 Notictis Ameghino, 1889, p. 911; Kraglievich, 1934, p. 62; Reig, 1952, p. 3. Type. — Notictis ortizi Ameghino, 1889, p. 912. Distribution. — "Mesopotamian beds" along rio Parana, Entre Rfos Province, Argentina. Diagnosis. — As for type and only known species. Notictis ortizi Ameghino, 1889. Figures 17, 18; Tables 2, 3. Notictis ortizi Ameghino, 1889, p. 912, pi. 72, fig. 14, pi. 81, fig. 7; 1891b, p. 262; Krag- lievich, 1934, pp. 62-63; Simpson, 1945, p. 42n (as nomen vanum); Reig, 1952, p. 3, fig. 1; Ringuelet, 1953, p. 267. Didelphys curvidens Burmeister, 1891, p. 379, pi. 7, fig. 1. Type. — MACN 3996, a left mandibular ramus with alveoli of C and F^-Mx complete, M2.4 present but broken (figured by Ameghino, 1889, pi. 72, fig. 14; Burmeister, 1891, pi. 7, fig. 1; Reig, 1952, fig. 1). Hypodigm. — Type only. Horizon and locality. — "Piso mesopotamico de la formacion patagonica," "bar- rancas de los alredores de la ciudad del Parana," along the rio Parana, Entre Rios Province, Argentina. Age. — Huayquerian. Fie. 17. Notictis ortizi Ameghino, 1889 (Huayquerian). MACN 39% (type), a left man- dibular ramus with alveoli of C, PpM! complete, and M2A present but broken: a, labial; b, occlusal; c, lingual views. Scale = 20 mm. f^^^W Fie. 18. Notictis ortizi Ame- ghino, 1889 (Huayquerian). Stereopairs of MACN 3996 (type), a left mandibular ra- mus with alveoli of C, P|-M, complete, and M2.4 present but broken: a, labial; b, occlu- sal; c, lingual views. Scale = 30 mm. 25 26 FIELDIANA: GEOLOGY Diagnosis. — Very small size, smallest known species of Huayquerian bor- hyaenid; C weakly developed; P, tiny and set obliquely in jaw; P23 larger and crowded, but roots aligned in same anteroposterior axis; P, separated from P2 by small diastema; P2 and P3 crowded such that posterior alveolus of former and anterior alveolus of latter are confluent; M^-, have small, but basined talonids, well-developed anterobasal cingula, and lingual side of talonid basin slightly higher than labial side; differs from Pseudonotictis pusillus in having relatively larger talonids, in an absolutely shorter premolar region, and smaller P,. Comments. — The type of Notictis ortizi has been redescribed and its taxonomic history discussed by Reig (1952). As demonstrated by Ameghino (1891b, p. 262) and Kraglievich (1934, pp. 62-63), Didelphys curvidens Burmeister (1891, p. 379) and N. ortizi Ameghino (1889, p. 912) were founded on the same specimen, MACN 3996. A second specimen, MACN 565 (cast) originally referred to N. ortizi and figured by Ameghino (1889, pi. 81, fig. 7) and later discussed by Reig (1952, p. 3), is almost certainly a didelphoid. Notictis ortizi very probably evolved from the Santacrucian species Pseudono- tictis pusillus. The probable phylogenetic relationships of these species are dis- cussed in the comments section on P. pusillus. Perathereutes Ameghino, 1891c Perathereutes Ameghino, 1891c, p. 313. Type. — Perathereutes pungens Ameghino, 1891c, p. 313. Distribution. — Santa Cruz Formation, Santa Cruz Province, Argentina. Diagnosis. — As for type and only known species. Perathereutes pungens Ameghino, 1891c. Figures 19-21; Table 4. Perathereutes pungens Ameghino, 1891c, p. 313; 1894, p. 392, fig. 54; 1898, pp. 191, 193, fig. 58d; 1935, p. 108, fig. 23 (caption only). Epanorthus aratae Ameghino, 1889, pi. 1, figs. 10-lOb (partim). Abderites altiramis Ameghino, 1894, p. 340; 1898, p. 186. Type of Perathereutes pungens. — MACN 684, a left mandibular ramus with al- veolus of C, roots of P13, MM complete but worn (figured by Ameghino, 1894, fig. 54; 1898, fig. 58d). Type of Abderites altiramis. — MACN 8250, a fragment of a right mandibular ramus with talonid of M3, anterior alveolus and posterior root of M4 (figured by Ameghino, 1889, pi. 1, figs. 10-10d as Epanorthus aratae). Hypodigm. — Types only. Horizon and locality. — Both types are from the Santa Cruz Formation, Santa Cruz Province, southern Argentina. MACN 684 is from Monte Observacion and was collected by C. Ameghino in 1890-91. MACN 8250 is from La Cueva and was collected by C. Ameghino in 1892-93. Age. — Santacrucian. Diagnosis. — Intermediate in size between Pseudonotictis pusillus and Sipalocyon gracilis; P] small and set at slight oblique angle in jaw and separated from C and P2 by small but distinct diastems; talonid well developed and basined on M,^, but relatively smaller than in Sipalocyon; mandibular ramus exceptionally shallow in depth relative to its length; symphysis extends posteriorly to point below P2.3 contact; C moderately developed as in Sipalocyon; entoconid higher than hypoconid on M^, lower on M4. u B — 5 I Cm " 3 a, a, 2 27 Fig. 20. Perathereutes pungens Ameghino, 1891c (Santacrucian). Stereopairs of MACN 684 (type), a left mandibular ramus with alveoli of C, roots of Pj.3, and Mw complete but worn: a, labial; b, occlusal; c, lingual views. Scale = 40 mm. A* A. ..'j**N*we\ - Tip«» fl.fc$4 Fig. 21. Perathereutes pungens Ameghino, 1891c (Santacrucian). MACN 684 (type), a left mandibular ramus with alveoli of C, roots of Pi.3, and M^ complete but worn: top, occlusal; bottom, lingual views. Scale = 50 mm. 28 E£ 2^ 0> 00 v- cu o T o uS in 00 vO zz uu 29 30 FIELDIANA: GEOLOGY Description. — Three mental foramina occur on type: largest below diastema between P, and P2, a slightly smaller one below P2_3 contact, and one of similar size to latter below Mi_2 contact. The region of the mandibular ramus near the P2 and anterior root of the P3 is covered with secondary bone growth, indicating that this area was either diseased or injured during the life of the animal. Comments. — Abderites altiramis was erected by Ameghino (1894, p. 304) on a fragment of a right mandibular ramus (MACN 8250) that preserves the anterior alveolus and talonid of M3 and roots of M4. Ameghino characterized this species as being almost twice as large as Abderites crassiramis. MACN 8250 is not, how- ever, an abderitine, but a small borhyaenid. Size of the M^ roots and structure of the M3 talonid agree almost perfectly with the type of Perathereutes pungens (MACN 684). The ramus of A. altiramis is slightly shallower and more gracile than in P. pungens, but these differences are minor, and A. altiramis is regarded a junior synonym of P. pungens. MACN 8250 was figured by Ameghino (1889, pi. 1, figs. 10-10b) as Epanorthus aratae. Ameghino (1891c, p. 313) originally placed P. pungens in the family Thylacynidae, but later (1894, p. 392) included it in his group Sparassodonta, family Amphiproviverridae. Perathereutes pungens is similar to Sipalocyon gracilis, although there are impor- tant differences; P. pungens differs from specimens of the latter (i.e., MACN 5938) in being smaller in size, in having Pj set at a sharper oblique angle in jaw and being separated from C and P2 by distinct diastems, and in a proportionately smaller talonid basin, especially on M4. Although isolated teeth and partial jaw fragments of these species may be difficult to separate, more complete material as described here shows these forms to be distinct and the taxa valid. The most likely ancestor for P. pungens is the Riochican species Patene simpsoni. These taxa are similar in size and structure, but differ in P. pungens having P, set obliquely in jaw and being separated from C and P2 by distinct diastems, in lacking a metaconid, in a slightly more reduced talonid on MM, and in a more elongated and tricuspid M,. Perathereutes pungens is in turn similar to Borhyaenidium musteloides from beds of Huayquerian age and most probably represents an ancestor of that species. Both species have Px set at a slight oblique angle in the jaw, the Px is separated from the C and P2 by distinct diastems, and talonids (especially on M4) are more reduced than in species of Sipalocyon. Borhyaenidium musteloides differs from P. pungens primarily in being slightly larger in size and in the talonid basin being proportionately and absolutely more reduced. The fact that talonid reduction is already begun in P. pungens makes this species a more likely ancestor for B. musteloides than does S. gracilis. For S. gracilis to have been ancestral to B. musteloides, it would have had to have undergone talonid reduction and hence passed through a stage similar to that seen in P. pungens. The S. gracilis lineage, however, with the larger talonid basins appears to have continued into the Montehermosan as evidenced by Notocynus hermo- sicus. The lineages that terminated in the reduced talonid form of B. musteloides in the Huayquerian and the large talonid form of N. hermosicus in the Monte- hermosan were thus already distinct in the Santa crucian. Borhyaenidium Pascual & Bocchino, 1963 Borhyaenidium Pascual & Bocchino, 1963, p. 101. Type. — Borhyaenidium musteloides Pascual & Bocchino, 1963, p. 101. MARSHALL: REVIEW OF THE HATHLYACYNINAE 31 Distribution. — Known from beds of Huayquerian and Montehermosan age in Argentina. Diagnosis. — Similar in size to Sipalocyon and Notocynus; a large mental foramen occurs below posterior root of P,; P, separated from C and P2 by small diastems; talonid distinct, not basined, and relatively smaller than in Sipalocyon and No- tocynus; trigonid relatively narrower and more elongated than in Notocynus, Sipalocyon, and Perathereutes, and paraconid lies more directly anteriad of pro- toconid; protocone distinct but smaller than in Sipalocyon, Pseudonotictis, and Cladosictis, and set more anteriad relative to para- and metacone as in Chasicostylus castroi; paracone small and fused basally to larger metacone and becoming smaller from M1 to M\ whereas metacone becomes larger in same direction; para- and metacone set more labiad on M1 and more mediad on M3; parastyle large and elongated anteriorly on M1 and in same axis as para- and metacone; parastyle on M: shorter and set on anterolabial corner of tooth labiad to para- metacone axis; parastyle on M3 smaller and set completely on labial side of tooth; M3 with small but distinct ectoflex opposite metacone; metacrista well developed and its shear surface continues along inner surface of protocone. Borhyaenidium musteloides Pascual & Bocchino, 1963. Figures 22-25; Tables 5, 6. Borhyaenidium musteloides Pascual & Bocchino, 1963, p. 101, pis. 1, 2. Type. — MLP 57-X-10-153, poorly preserved partial skull with left M1'3 and right M2"3; a nearly complete left mandibular ramus with C, roots of P„ P23 complete, posterior half of M„ M; complete, and Mw missing tip of protoconid; a fragment of a right mandibular ramus with complete P3-M, and broken M:^; proximal and distal ends of a humerus; proximal end of an ulna; and assorted postcranial fragments, all of a single associated individual (figured by Pascual & Bocchino, 1963, pis. 1, 2). Hypodigm. — Type only. Horizon and locality. — Type from Epecuen Formation, Salinas Grandes de Hi- dalgo, La Pampa Province, Argentina. Age. — Huayquerian . Diagnosis. — Differs from Borhyaenidium riggsi in having, in most cases, slightly smaller linear tooth dimensions (see table 5); a slightly smaller talonid on M4; and a slightly smaller paracone (especially on M3). Comments. — It is probable that Perathereutes pungens is the Santacrucian ances- tor of Borhyaenidium musteloides. Perathereutes pungens is slightly smaller than B. musteloides, and its talonids are slightly larger and more distinctly basined. How- ever, these taxa are similar in having P, separated from C and P2 by small but distinct diastems. In Pseudonotictis pusillus the talonids are proportionately smaller relative to trigonids than occurs in species of Sipalocyon and Cladosictis. We thus see in P. pusillus the incipient talonid reduction that is continued in this lineage and that appears to a more extreme degree in B. musteloides. Borhyaenidium musteloides has a dental structure very similar to that of Chasi- costylus castroi. Both species have the upper molars greatly elongated antero- posteriorly, the parastyle on M1 is very large and projects anteriorly in the same axis as para- and metacone, whereas on M2 the parastyle is shorter and is set at the anterolabial corner of the tooth labiad to the para-metacone axis; protocone is reduced and is set anteriad of para- and metacone, and shear along metacrista 32 FIELDIANA: GEOLOGY Fig. 22. Borhyaenidium musteloides Pascual & Bocchino, 1963 (Huayquerian). MLP 57-X- 10-153 (type), a left maxillary with M1"3: a, labial; b, occlusal; c, lingual views. Scale = 10 is increased by elongation of that area and by its extension anteriorly along inner surface of protocone; lower molars are also elongated anteroposteriorly, espe- cially the protoconid and paraconid, and the talonid is more reduced than occurs in species of Sipalocyon, Cladosictis, and Notocynus. These features are thus very distinct, and they do not occur jointly in other borhyaenid species of similar size. Chasicostylus castroi differs from B. musteloides in being larger in size, in having a relatively larger protocone and talonid basin, and in having the protocone not set as far anteriad relative to the para- and metacones. Apart from these few differences, these taxa are structurally very similar. The dental specializations seen in C. castroi have been carried further in B. musteloides, in which the protocone is more reduced and is set more anteriad of the para- and metacone, resulting in elongation of the metacrista shear by making it more continuous and unobstructed. Borhyaenidium musteloides thus has slightly more advanced carnassial specializations than does C. castroi. z- o 9 i SB CM rl O ~ -r ^* «-l © o -r «* in r*S cS •*! P o 49 •t in in v© in vC On O * ao oo 00 f*i Tf ^1" ni cj »N . o o "♦ •** °) : no vO in in in ■* hx ^f in _ tn ro rj ni cj \o m in in v© in 33 I I I Fig. 23. Borhyaenidium musteloides Pascual & Bocchino, 1963 (Huayquerian). Stereopairs of MLP 57-X-10-153 (type), a left maxillary with M1"3: a, labial; b, occlusal; c, lingual views. Scale = 10 mm. Table 6. Measurements of mandibular rami of Borhyaenidium musteloides, B. riggsi sp. nov., and Notocynus hermosicus. Depth of Depth of ramus below ramus below labial side Breadth of labial side Breadth of Specimen of M, same of M4 sami Borhyaenidium musteloides MLP 57-X-10-153(l) MLP 57-X-10-153(r) 11.0 11.0 4.3 4.4 11.8 11.7 4.9 5.0 Borhyaenidium riggsi sp. nov. FMNH P14409(l) FMNH P14409(r) 11.0 10.9 4.2 3.8 10.2 9.9 5.4 5.2 Notoa/nus hermosicus MLP 11-91 10.7 4.2 34 £S 35 Fig. 25. Borhyaenidium musteloides Pascual & Bocchino, 1963 (Huayquerian). Stereopairs of MLP 57-X-10-153 (type), a nearly complete left mandibular ramus with C, roots of Pu P2.3 complete, posterior half of M,, M2 complete, and M3^ missing tip of protoconid: a, labial; b, occlusal; c, lingual views. Scale = 50 mm. 36 MARSHALL: REVIEW OF THE HATHLYACYNINAE 37 Borhyaenidium riggsi sp. nov. Figures 26-29; Tables 5, 6. Notocynus hermosicus Riggs & Patterson, 1939, p. 149. Etymology. — riggsi named in honor of Elmer S. Riggs, leader of the 1926-1927 Second Marshall Field Paleontological Expedition to Catamarca Province, Ar- gentina. Type. — FMNH P14409, two fragments of left maxillary, one with C and anterior root of P' and the other with M1'1; a left mandibular ramus with root of C, roots of P,.„ M, almost complete but missing anterior edge, M: missing talonid, and Mw complete; and a broken right mandibular ramus with roots of C-M4; all of a single, associated individual. c3o<a^^ Fig. 26. Borhyaenidium riggsi sp. nov. (Montehermosan). FMNH P14409 (type), a left mandibular ramus with root of C, roots of P^j, M! almost complete but missing anterior edge, M2 missing talonid, and Mw nearly complete: a, labial; b, occlusal; c, lingual views. Scale = 30 mm. Hypodigm. — Type only. Horizon and locality. — Type collected by R. D. Thorne from unit 32 of Corral Quemado Formation at Puerta de Corral Quemado, Departamento de Bel£n, Catamarca Province, Northwest Argentina. Age. — Montehermosan. Diagnosis. — Differs from B. musteloides in having, in most cases, slightly larger linear tooth dimensions (see table 5); a slightly larger talonid on M4; and a slightly larger paracone (especially on M}). Fig. 27. Borhyaenidium riggsi sp. nov. (Montehermosan). Stereopairs of FMNH P14409 (type), a left mandibular ramus with root of C, roots of P^, M, almost complete but missing anterior edge, M2 missing talonid, and Mw nearly complete: a, labial; b, occlusal; c, lingual views. Scale = 40 mm. 38 MARSHALL: REVIEW OF THE HATHLYACYNINAE 39 Comments. — The type of B. riggsi is the specimen referred by Riggs & Patterson (1939, p. 149) to Notocynus hermosicus. This assignment was made largely on the basis of the similarity in size of these animals and their occurrence in beds of similar age. However, these specimens differ in N. hermosicus having a more crowded premolar series with no diastems separating P, from C or from P2, and in the M3 talonid being relatively larger and distinctly basined. Borhyaenidium riggsi is tentatively recognized as a species distinct from B. musteloides. The three characters I have used to distinguish these species may yet be shown to represent only individual variation within members of a single species. Both species are known only by their types, which come from beds of different ages as demonstrated by study of the associated faunas. These spec- imens are clearly similar, and the species may be regarded as representing a single evolutionary lineage. I therefore recognize B. musteloides as the Huay- querian ancestor of B. riggsi. Fie. 28. Borhyaenidium riggsi sp. nov. (Montehermosan). FMNH P14409 (type), two frag- ments of a left maxillary, one with C and anterior root of P' and other with M1 3: a, labial; b, occlusal; c, lingual views. Scale = 20 mm. 40 FIELDIANA: GEOLOGY Fig. 29. Borhyaenidium riggsi sp. nov. (Montehermosan). Stereopairs of FMNH P14409 (type), two fragments of a left max- illary, one with C and anterior root of P1 and other with M1 3: a, labial; b, occlusal; c, lingual views. Scale = 30 mm. Sipalocyon Ameghino, 1887 Sipalocyon Ameghino, 1887, p. 8; 1889, p. 292. Amphithereutes Ameghino, 1935, p. 108. Thylacodictis Mercerat, 1891a, p. 54. Protoproviverra Ameghino, 1891c, p. 312, nee Lemoine, 1891, p. 379 (Creodonta). Amphiproviverra Ameghino, 1891e, p. 397n, to replace Protoproviverra, which was preoc- cupied. Type of Sipalocyon. — Sipalocyon gracilis Ameghino, 1887, p. 8. Type of Amphithereutes. — Amphithereutes amputans (Ameghino, 1891c, p. 313). Type of Thylacodictis. — Thylacodictis exilis Mercerat, 1891a, p. 54. Type of Protoproviverra. — Protoproviverra manzaniana Ameghino, 1891c, p. 312. Type of Amphiproviverra. — Amphiproviverra manzaniana (Ameghino, 1891c, p. 312). Distribution. — Colhuehuapian and Santacrucian, Patagonia, southern Argen- tina. MARSHALL: REVIEW OF THE HATHLYACYNINAE 41 Diagnosis. — Small to medium-sized borhyaenids; mandibular rami gracile and shallow in depth; lower premolars increase in size from P, to P,; all have distinct posterobasal cusps or heels that increase in size from P, to P,; P, either separated from P: and C by tiny diastems or all abut (character variable), when tightly packed anterior roots tend to lie slightly labiad of posterior roots; lower molars increase slightly in size from M, to M4, all have a very large and deeply basined talonid, and distinct anterobasal cingula; protoconid usually twice as large as paraconid; lingual side of talonid basin higher than labial side in unworn teeth; IM subequal in size but with slight decrease from I1 to P; upper premolars increase in size from P1 to P3; P1 usually separated from C and P: by small diastema; P: may or may not be separated from P3 by small diastema; P1 3 aligned in same anteroposterior axis; M1"4 with large protocone and deeply basined trigonid; distinct para- and metaconules; paracone and metacone fused basally; paracone becomes smaller relative to metacone from M1 to M3; metacone becomes relatively Fie. 30. Sipaloqfon externa (Ameghino, 1902c) (Colhuehuapian). MACN 52-383 (type), left F°-M3: a, labial; b, occlusal; c, lingual views. Scale = 20 mm. 42 FIELDIANA: GEOLOGY and absolutely larger from M1 to M3; M4 with large protocone and paracone, metacone absent; weak but distinct parastyle on M1'3 which becomes larger from M1 to M3; parastyle on M4 connected with paracone by parastylar ridge; M3 has shallow but distinct ectoflex; infraorbital canal large and opens over anterior root of P3; supraorbital processes weakly developed. Sipalocyon externa (Ameghino, 1902c). Figures 30, 31; Table 7. Cladosictis externa Ameghino, 1902c, p. 129. Type. — MACN 52-383, a partial cranium (poorly preserved) with left P3-M3, and roots of right P1, P2-M2, and complete M4. Hypodigm. — Type only. Horizon and locality. — Colhue-Huapi Formation, Chubut Province, Argentina; specimen labeled "la barranca al sur del Lago Colhue-Huapi"; collected by C. Ameghino. Age. — Colhuehuapian. Diagnosis. — Slightly smaller than specimens of Sipalocyon gracilis. Comments. — Sipalocyon externa is slightly smaller and relatively more gracile than specimens of S. gracilis from the Santacrucian. Morphologically, these spe- cies are very similar. Compared with some specimens of S. gracilis (e.g., MACN %i*-i% Fig. 31. Sipalocyon externa (Ameghino, 1902c) (Col- huehuapian). Stereopairs of MACN 52-383 (type), left I^-M3: a, labial; b, occlusal; c, lingual views. Scale = 20 mm. in . v© : a. ^ 3 CNj .-J in in m CM v© vC o o O ON vd in O ~ O O 43 44 FIELDIANA: GEOLOGY 5959), S. externa has a slightly smaller protocone, a proportionately shorter meta- crista, and a more distinct ectoflex on the M3 and a proportionately larger one on the M2. There is sufficient variation, however, in the large comparative sample of S. gracilis from the Santacrucian to include a significant part of these differences and to cast doubt on the validity of these features as diagnostic characters. The slightly smaller size of MACN 52-383 appears to be the only nondisputable character that can be used to separate it from specimens of S. gracilis. In view of their close similarity in size and morphology, S. externa is regarded as the Colhuehuapian ancestor of S. gracilis. Sipalocyon gracilis Ameghino, 1887. Figures 32-43; Tables 8-13. Sipalocyon gracilis Ameghino, 1887, p. 8; 1889, p. 292; 1891c, p. 315; 1891d, p. 354; 1894, pp. 393, 394, fig. 55; 1898, pp. 191, 193, fig. 58e; 1935, p. 110, fig. 25 (caption only). Protoproviverra manzaniana Ameghino, 1891c, p. 312. Amphiproviverra manzaniana Ameghino, 1894, p. 389, fig. 52; 1898, p. 193, figs. 58a, b; 1906, fig. 188; 1935, p. 107, figs. 4, 27 (captions only); Sinclair, 1906, p. 400 (with numerous plates and figures); Schlosser, 1925, p. 35, fig. 56; Scott, 1937, fig. 415. Thylacodictis manzaniana Cabrera, 1927, p. 304, fig. 17B. Perathereutes amputans Ameghino, 1891c, p. 313; 1894, p. 393. Amphithereutes amputans Ameghino, 1935, p. 108, fig. 33 (caption only). Perathereutes obtusus Ameghino, 1891c, p. 313; 1894, p. 392; 1935, p. 108, fig. 7 (caption only). Thylacodictis obusta Cabrera, 1927, p. 302, figs. 15, 17c. Hathliacynus lynchi Mercerat, 1891a, p. 53; Ameghino, 1894, p. 387 (as junior synonym of Cladosictis patagonica). Thylacodictis exilis Mercerat, 1891a, p. 54; Ameghino, 1891d, p. 354 (as junior synonym of S. gracilis); Cabrera, 1927, p. 303, figs. 16, 17. Amphiproviverra minuta Ameghino, 1894, p. 390; 1935, p. 107, fig. 6 (caption only); Sinclair, 1906, p. 404, pi. 59, figs. 3, 3a, pi. 60, figs. 3, 3a. Sipalocyon curtus Ameghino, 1894, p. 394; 1935, p. Ill, fig. 28 (caption only). Sipalocyon mixtus Ameghino, 1894, p. 394; 1935, p. Ill, fig. 11 (caption only). Amphiproviverra crassa Ameghino, 1894, p. 391. Thylacodictis crassa Simpson, 1930, p. 50. Sipalocyon altiramis Ameghino, 1894, p. 395. Sipalocyon longus Ameghino, 1894, p. 395. Type of Sipalocyon gracilis. — MACN 647, a fragment of a right mandibular ramus with alveoli of I^ root of C, alveoli of Pu P2-M2 complete (figured by Ameghino, 1894, fig. 55; 1898, fig. 58e). Type of Thylacodictis exilis. — MLP 11-12, a fragment of a right mandibular ramus with alveolus of C, roots of P,.3, M! complete, and anterior half of M2. Type of Amphiproviverra minuta. — MACN 685, part of an isolated left M2 with proto-, para-, and metacones, but lacking metastylar region (protocone was at one time broken off and has since been glued back on the specimen). Type of Protoproviverra manzaniana. — MACN 691-703, a left and right mandib- ular ramus with complete dentition (691), a left maxillary and premaxillary with dentition (692), and associated fragmentary postcranial remains (693-703) (fig- ured by Ameghino, 1894, fig. 52; 1898, figs. 58a, b; 1906, fig. 188; listed as "tipo" in Ameghino's catalogue and fits original description perfectly). Type of Perathereutes amputans. — MACN 682, a fragment of a right mandibular ramus with alveoli of C-P2, roots of P3, M,.2 complete but worn, anterior half and posterior root of M3 (listed as "tipo" in Ameghino's catalogue and matches original description perfectly). •* Fig. 32. Sipalocyon gracilis Ameghino, 1887 (Santacrucian). MACN 647 (type), a fragment of a right mandibular ramus with alveoli of Ij.3, root of C, alveoli of P,, P2-M2 complete: top, occlusal; bottom, labial views. Scale = 30 mm. T< • ** . • ~' - ft&B&S 4 Fie. 33. Sipalocyon gracilis Ameghino, 1887 (Santacrucian). MACN 5965 (type of Sipalocyon longus), a right mandibular ramus with alveolus of C, roots of P, and M4, and P2-M3 complete: top, occlusal; bottom, lingual views. Scale = 50 mm. 45 Fig. 34. Sipalocyon gracilis Ameghino, 1887 (Santacrucian). MACN 5963 (type of Sipalocyon mixtus), a fragment of a right mandibular ramus with roots of C, Pu P3 and M2/ and with P2, Mi, and M3 complete: top, occlusal: bottom, labial views. Scale = 30 mm. Fig. 35. Sipalocyon gracilis Ameghino, 1887 (Santacrucian). MACN 5938, a nearly complete left mandibular ramus with dentition: top, occlusal; bottom, lingual views. Scale = 50 46 MARSHALL: REVIEW OF THE HATHLYACYNINAE 47 Fie. 36. Sipalocyon gracilis Ameghino, 1887 (Santacrucian). MACN 5964 (type of Sipalocyon altiramis), greater part of a right mandibular ramus with PpM* complete: top, occlusal; bottom, lingual views. Scale =50 mm. Type of Perathereutes obtusus. — MACN 683, a fragment of a right mandibular ramus with root of C, alveoli of P„ roots of P23, M, complete, and anterior half of M2 (listed as "tipo" in Ameghino's catalogue and matches original description perfectly). Type of Sipalocyon altiramis. — MACN 5964, a right mandibular ramus with P,- M< complete (not listed as type in Ameghino's catalogue or on specimen, but original description fits perfectly). Type of Sipalocyon curtus. — MACN 5960-62; 5960, a fragment of a right man- dibular ramus with C, posterior half of Pu P2 complete, and anterior root of P3; 5961, assorted tooth fragments (both upper and lower); and 5962, a caudal vertebra (all of a single individual). Type of Sipalocyon longus. — MACN 5965, a right mandibular ramus with al- veolus of C, roots of P, and M^, and P2-M3 complete (not listed as type in Ameghino's catalogue or on specimen, but it is only specimen listed by this name in his catalogue, and original description fits perfectly). Type of Sipalocyon mixtus. — MACN 5963, a fragment of a right mandibular ramus with roots of C, P„ P3, and M2, and with P2, M,, and M3 complete. Syntypes of Amphiproviverra crassa. — MACN 5957, a left mandibular ramus with roots of P„ P2.3 complete, M,.3 broken, M4 complete; and MACN 5958, a left maxillary fragment with P'-M4 complete. (These specimens have slightly different preservation, and the teeth in MACN 5958 are more deeply worn than in MACN 5957, indicating that they are of different individuals, but are certainly of the same species. They are not listed as types in the catalogue or on the specimens, but they are the only specimens listed under the name A. crassa in Ameghino's catalogue, and they fit the original description of this species perfectly.) 48 MARSHALL: REVIEW OF THE HATHLYACYNINAE 49 Table 8. Measurements of mandibular rami of Sipalocyon gracilis. Depth of ramus Depth of ramus Specimen below labial Breadth of below labial Breadth of side of M, same side of Mi same MACN 647 10.5 4.6 ... ... MACN 668 10.7 5.0 ... ... MACN 682 12.3 5.3 ... ... MACN 691(1) 5.1 5.3 MACN 691 (r) 12.6 5.3 13.2 5.7 MACN 5938 10.5 5.0 11.8 5.2 MACN 5940 10.7 5.0 11.6 5.3 MACN 5953 13.7 6.3 MACN 5963 13.7 5.1 ... MACN 5964 13.6 5.6 14.6 5.5 MACN 5965 12.8 4.9 13.3 5.2 MACN 9353 11.5 5.0 ... MLP 11-12 11.2 5.1 ... MLP 11-41 12.6 5.2 ... Table 9. Statistics for some dimensions of mandibular rami of Sipalocyon gracilis. Dimension N OR X s cv Depth of ramus below labial side of M, 12 10.5-13.7 11.89 1.19 10.01 Breadth of ramus below M| 13 4.6-5.6 5.09 0.24 4.72 Depth of ramus below labial side of M ; 6 11.6-14.6 13.03 1.15 8.83 Breadth of ramus below M 7 5.2-6.3 5.50 0.40 7.27 Lectotype of Hathliacynus lynchi. — MLP 11-7, a fragment of a left mandibular ramus with M,_, complete. Type of Thylacodictis exilis. — MLP 11-12, a fragment of a right mandibular ramus with alveoli of C-P„ roots of P23, M, complete, anterior half of M, complete. Hypodigm. — The 13 types and MACN 648, a fragment of a right mandibular ramus with alveoli of P12, and P3 complete; MACN 667, a fragment of a right mandibular ramus with C-P2 complete, and alveoli of P^ MACN 668, a fragment of a left mandibular ramus with posterior root of P2, roots of P3, and M13 complete; MACN 688, a right maxillary fragment with root of C, P1 complete, and alveoli of P23 (listed as Amphiproviverra manzaniana in Ameghino's catalogue); MACN 704, a left maxillary fragment with M2'3 (listed as A. manzaniana in Ameghino's catalogue); MACN 5938-5949, associated left and right mandibular rami almost complete, a maxillary fragment with left M2~\ right M3-1, crowns of both upper C's, and various postcranial elements, all of a single individual; MACN 5952, greater part of a cranium with most of dentition; MACN 5953, a fragment of a left mandibular ramus with Mw (same individual as MACN 5952); MACN 5959, rostral portion of a skull with most of dentition of left side and alveoli of IM, root of C, and P1 complete on right side; MACN 6286, a fragment of a left mandibular ramus with posterior alveolus of P, and P23 (listed as Anodonictis obliterata in Ameghino's catalogue); MACN 9352, a left maxillary fragment with alveolus of C, PM complete, P3-M2 broken, alveoli of M5"4 (listed as A. manzaniana in Ameghino's catalogue); MACN 9353, a fragment of a left mandibular ramus ONCTvCviCT^ ii-icvj : ni 06 06 ov ov i-"»-"(N>-l -fS(N • CM I-H .-l ,-H r-l NOvvOO t vo m m . ■* ■* \C \D MTl Ov t}< rj< in •<}< CM CM CM CM CM q q o q o r-« \0 >C «C \0 N >C in in in q i-h ts in in vd vd « S •I- o s & rx 3 vO [^ r^ 3C o ts OV 00 ON X, r^ r-v O <-H vd vO in in «' vC in t< ts t< iri in vd vd in ts CM in in m vO in in CM q CN CD K \0 vc sC ^c ^c ts vd K t< cn ^c sC vd vd O q in 00 ** oo q X Ov Tt in in vd ** ^ in 15 \6 -* >* in in o co in nO in TT vO in vO •£: co to oo ts ts is! nC sd ^c CN vd K N K sC 43 vd vd i tS. °°. m oo . ■«* in °°. 3C CM CM •<* co t*5 >* CO 1 Tf •* m rn ■* ■<* q -* co (N •* in m ^r en q ts vd d vO l< vO K N vd sC vd vd in ov in CM CM CM VOH^f in vd in i— i in in in vo CO CM CM CM CM T* T}< CM VO tS vd in in in in . . oo . . . . ts Ov 00 . 00 o . . 00 O O Ov • ,— i • I— 1 I— 1 T— 1 • t-H CM • •HMNH . . o . . . . Ov CM O .H(V| . O CM O O in in m in i i ts in in ts oo in in vo ts • oo o : CO •<* 00 00 CM Ov Ov 1 CO CO 1 CO CO 1 Ov 00 Ov i— i O CO CO CO "^ I1 S-i^Si- Si- ts ts. OvOvCMCM00OvOvCMts.ts inoocMT£cocoinininininin<oco 0000OvOOvOv0vOv0v0vOvC0'rl'rl_-, vovovotsinininininininovCiHCuoo ZZZZZZZZZZZZXXK UUUUUUUUUUUU22rt <<<<<<<<<<<<5§r) 50 MARSHALL: REVIEW OF THE HATHLYACYNINAE 51 Table 11. Statistics for some upper cheek tooth dimensions of Sipaltxyon gracilis. N OR x s CV Dimensions P1 L W P2 L W P3 L W M1 L W M2 L W MJ L W M4 L W P'-M3 M'-3 L L 13 3.8-4.4 3.% 0.18 4.55 13 1.4-1.8 1.58 0.12 7.59 10 4.9-5.2 5.06 0.11 2.17 10 1.7-2.0 1.87 0.11 5.88 10 5.2-6.4 5.73 0.39 6.81 10 2.5-3.2 2.71 0.27 9.96 11 6.3-7.5 6.74 0.46 6.82 11 3.7-4.5 4.10 0.32 7.80 14 6.3-7.6 6.91 0.48 6.95 12 4.5-6.0 5.16 0.49 9.50 14 6.0-7.5 6.69 0.52 7.77 14 5.7-7.9 6.46 0.82 12.69 12 5.5-7.5 6.25 0.65 10.40 12 2.4-2.9 2.55 0.17 6.67 10 36.2^13.3 38.10 2.99 7.85 12 18.5-22.6 20.28 1.41 6.95 with posterior root of P2, Pj-M, complete, roots of M^ MACN 9354, a fragment of a right mandibular ramus with P, complete, roots of P2, P3-M, complete; MACN 9355, maxillary fragment with a molar; MACN 9356, maxillary fragment with a molar; MACN 9357, an isolated C; MACN 9358, distal end of a humerus (MACN 9353-9358 are of a single individual); MACN 9389, a fragment of a left mandibular ramus with roots of P,.2, and P3-M2 complete, M3 missing tips of protocone and paracone; MACN 9390, an isolated left M1; MACN 9391, an iso- lated right M3; MACN 9392, an isolated vertebra and assorted tooth fragments (MACN 9389-92 are of the same individual, labeled Sipalocyon altiramis in Ame- ghino's catalogue); MLP 11-22, a fragment of a left mandibular ramus with roots of P,.2/ P3-M2 complete, anterior root of M^ MLP 11-25, a fragment of a right mandibular ramus with alveoli of P,, roots of P2, P3 complete, alveoli of M„ and MM complete; MLP 11-39, a fragment of a left mandibular ramus of a juvenile with M„ and associated fragment of a right mandibular ramus with P2 erupting (anterior edge is up, but posterior edge is not), alveoli of DP3, M, complete, and alveoli of M2 (both specimens of same individual) (no trace of P, because this portion of the ramus is broken away) (see Marshall, 1976a, 1978); MLP 11-41, a fragment of a left mandibular ramus with root of C, P, complete, roots of P2, P3 complete, and alveoli of M,; MLP 11-76, a fragment of a left mandibular ramus with M2; MLP 11-77, a fragment of a right mandibular ramus with M3; MLP 11- 105, an isolated C, and a right M,2, supposedly of same individual; FMNH P13777 [ = P13275J, greater part of skull (partially restored with plaster) with posterior half of left C, P', P3, M1'2 and M4 broken, and P2 and M3 complete, and roots of right P'-M4 (all broken), and three vertebrae in matrix; AMNH 9238, a fragment of a left mandibular ramus with M,.2 complete and M, broken; AMNH 9254, a nearly complete skull and partial mandibular ramus; PU 15029, an in- complete skull and mandible; PU 15148, a fragment of a right maxillary with alveoli of C, P1 complete, roots of P2, and P3-M4 complete; PU 15154, a crushed skull, atlas, third cervical, and portions of right fore- and left hindlimbs; PU 15312, fragments of lower jaws and upper teeth (exchanged to AC in 1922); PU 15373, an incomplete skull and mandible, fragments of a left ulna, phalanges, distal end of a metapodial, and part of a patella. ' /*4I*4 fy t /*▼ Y Fig. 38. Sipalocyon gracilis Ameghino, 1887 (Santacrucian). Stereopairs of MACN 691 (type of Protoproviverra manzaniana), a nearly complete left mandibular ramus with complete dentition: a, labial; b, occlusal; c, lingual views. Scale = 50 mm. 52 5 ! J »ON00 O in >d ni csj : vd H «n <s H ■ <N 3 O O «s cs >© f-j ^* ^* co co co ^" OOf TMfiN N ts >C >0 vO N ^* in 3 3 8 33 00 CM SS5 co co co co ■<* co 1 S I o 1 I = Ou 3 B « s 2 : <n o o e*i rj -r • >£> -T X t^ x T fi " co « -r *) n w " co ro JfJ m l«j en M ■ vO B> 00 M T . a* N CO at ao ir, r4 • vO c X C X • *© X X x x X x 9 . O Tf -T T o O CO c O ir, rj „ O 3~ rj • CO m t«J PI M C»J co t*J rt m t»J en en rj _ . CO ^r ir, T O a v© o o ir, T o _ O X • >© -c x £ £ x sD x x x X « £ nC tn . co r^ in o 00 CM CM . in -r sO 90 X in co co tn m ir, rj • CM ri cm rj rj rj r> ■ CM rj rj (M rj CM rj ri CI rj rj ir, . r^ _ o O O T ir, . PH <* x _ „ 90 rs r ^ X 9t 3~ ir, • in X x C x ir, ir, • >© ir, ir, x x ir, in ir, m m ID m ir. ir, T co cm in -f in -r C*J CO rj rt rj n rj r) CM rj CM fS cm ri ri rj ri n n 9 O 9 ir, m r^ B* 90 x ts cm ts C ir. in x ir. ir, ir, vO ir, ir, ir, ir, X tr, ir, ; m ^ ^ . in Tf «n ■ sO sO vO (SONh m <n co co <n o rsi <n vd sd sd vd 00 f> vO t~s CN <N <s <N Ninoos \d in in in ro <S in \o in in °. : *1 i^^R^ :^ric?a^c? : : : : : °^ P <n :f\i :fsirsirvicvi :c>ifvi<Ni-<rsi :»-<<n o .in . • oo oo o . . x oin fh(ni i*^^ iri :in • -ininin 'irivdininin • • • 'inin . .oo . . .ooooao ■ ■ . ts .oo • x. x '■ ~ :fsirsi«-<<-< • • tH :i-< :rHt-< ■ NClrnq • ^< ^< Tf» ^* ■ fS CM CM -r Tf co ^* in co ^* x x x ir, ir, O^ &* O* CO CO UUUUUUUUUUUUUUUUUUUfi.euti.DuO.ft.fi.ft.li.B.? ^ ^ ^ ^ ^ ^ ^ ^ ^- ^- ^ ^ ^. ^ ^ ^- ^ ^. ^ ^ ^ ^ ^ ^ ^ ^. x. ^ ; ^. < E C_ 53 54 FIELDIANA: GEOLOGY Table 13. Statistics for some lower cheek tooth dimensions of Sipalocyon gracilis. Dimension P. L W P2 L W Pa L W M, L W M2 L W M3 L W H, L W P1-M4 L MM L N OR CV 9 4.0-4.4 4.20 0.11 2.62 9 1.6-2.0 1.79 0.15 8.38 12 5.0-6.0 5.43 0.35 6.45 13 1.9-2.2 2.05 0.11 5.37 17 5.3-6.7 5.80 0.33 5.69 17 2.2-2.5 2.34 0.11 4.70 23 5.4-6.2 5.83 0.23 3.95 23 2.2-2.9 2.51 0.22 8.76 20 6.0-6.6 6.23 0.21 3.37 20 2.9-3.5 3.15 0.19 6.03 16 6.2-6.9 6.52 0.26 3.99 17 3.2-4.0 3.50 0.24 6.86 11 6.3-7.2 6.58 0.32 4.86 11 3.2-4.1 3.54 0.34 9.60 9 38.8-44.9 41.41 2.49 6.01 11 22.7-26.0 24.30 1.34 5.51 Horizon and locality. — All specimens are from the Santa Cruz Formation, Santa Cruz Province, Argentina. The specific localities of collection are as follows: Rio Santa Cruz MLP 11-7, 11-25; Monte Observacion MACN 647, 667, 668, 683, 685, 688, 704 (collected by C. Ameghino 1890-91), MACN 682, 6286 (collected by C. Ameghino 1891-92); Monte Leon MLP 11-12, 11-22, 11-39; La Cueva MACN 5938-5949, 5959, 5963, 5964, 5965, 9353, 9354 (collected by C. Ameghino 1892-93); Corriquen Kaik MACN 5952, 5953, 5957, 5958, 9352 (collected by C. Ameghino 1892-93); Shehuen ( = Sehuen) MACN 648, 691-703 (collected by C. Ameghino 1890-91); Yequa Quemada MACN 5960, 9389-91 (collected by C. Ameghino 1892-93); Killik Aike Norte ( = Felton's Estancia) PU 15373 (collected by J. B. Hatcher), AMNH 9254 (collected by B. Brown); La Costa, 7 miles south of Puerto Coyle FMNH P13777 (collected by F. G. Sternberg 1923); Canadon de las Vacas Fig. 39. Sipalocyon gracilis Ameghino, 1887 (Santacrucian). MACN 691 (type of Protopro- viverra manzaniana), a nearly complete right mandibular ramus with complete dentition: top, occlusal; bottom, labial views. Scale = 50 mm. I B C a XJ c E -3 E 2 E » SP" C > fl .R-.S ai 55 s* BWB)1 m ^rrl Fig. 41. Sipaloa/on gracilis Ameghino, 1887 (Santacrucian). Stereopairs of PU 15373, a nearly complete left mandibular ramus with most of dentition: a, labial; b, occlusal; c, lingual views. Scale = 50 mm. 56 MARSHALL: REVIEW OF THE HATHLYACYNINAE 57 AMNH 9238 (collected by B. Brown 1899); 10 miles south of Coy Inlet PU 15029 (collected by J. B. Hatcher 18%), 15148, 15154, 15312 (collected bv O. A. Peterson 18%); No data MLP 11-41, 11-76, 11-77, 11-105. Age. — Santacrucian. Diagnosis. — Slightly larger than Sipalocyon externa in linear tooth dimensions; differs from S. obusta in having a deeper and more robust mandibular ramus and a larger talonid on M . Description. — In the type of Sipalocyon gracilis, MACN 647, a very large mental foramen occurs below the P12 contact (its size is accentuated by the pathological disorder discussed below), and four smaller foramina occur more posteriorly — one below the middle of P2, one below P2.3 contact, one below the middle of Pj, and the largest of the four is below M,. The teeth are relatively unworn, indicating that this was a young animal. The PM are separated from each other by tiny diastems, and the alveolus of the anterior root of P, lies slightly labiad of the posterior alveolus. The P23 have small but distinct posterobasal cusps. A small but distinct entoconid and hypoconulid are easily distinguished on the talonid of M2, and together these cusps are about equal in size to the hypoconid. Periosteal proliferations. — Several specimens of S. gracilis (including MACN 647, 682, 683, 5%0, 6286) have a pathological disorder in the anterior part of the mandibular ramus. This infection is expressed as irregular rugosities around the mental foramina below the premolars and canine, often extending onto the symphyseal surface. In MACN 647, the type of S. gracilis, the bone in the symphyseal region and around the large mental foramen below the P,2 contact is swollen and is covered by numerous rugosities. The infection has resulted in a significant increase in size of this foramen. In MACN 682, the type of Amphithereutes amputans, a small rugosity occurs on the ventral border of the ramus below the P3-M, contact. In the type of Pera- thereutes obtusus, MACN 683, the symphyseal region and area below the pre- molars on the labial side is swollen and covered with rugosities. The anterior root of the P, is all but obliterated and is filled with secondary bone growth. This condition accentuated the size of the diastema between the C and P„ a feature noted by Ameghino in the original description of this species. In MACN 5960 the incisor region is greatly distorted by the infection and is an amorphous mass of secondary bone growth. The mental foramen below the P,.2 contact is extremely large and is surrounded and filled by rugosities. The ventral border of the ramus and the inner side of the symphysis are also affected, and rugosities extend to a point below the anterior root of the Pj. The C is very blunt, suggesting that infection caused malocclusion. The most extreme case of the infection occurs in MACN 6286. The entire canine alveolus and anterior root of the P, are filled with cancellous bone, and it appears that these teeth were either lost during the life of the animal or were never present. The ventral border of the symphysis is also very swollen, and the mental foramen below the P,.2 contact is enormous (4x2 mm), and the border is very swollen. Comments. — MACN 648 is listed as Sipalocyon gracilis in Ameghino's catalogue and on the specimen. There is, however, no direct reference to this specimen in Ameghino's description of this species, and it therefore does not appear to represent a syntype. MACN 648 is slightly larger than the type, MACN 647, and 58 FIELDIANA: GEOLOGY the posterobasal heel on the P3 is better developed. The Pj is separated from P2 by a small (1.5 mm) diastema. MACN 667 is listed as the "tipo" of Sipaloa/on mixtus in Ameghino's catalogue. However, he neither gave measurements for nor made special reference to this specimen in the original description of this species. The specimen used by Ameghino in the original description of S. mixtus is clearly MACN 5963, and this is the type. MACN 5963 is listed as S. mixtus in Ameghino's catalogue, but is not listed as the "tipo" either in the catalogue or on the specimen. The incisor alveoli of this specimen are broken away, and these teeth are not "rudimentary" as believed by Ameghino. In Ameghino's catalogue MACN 668 is listed as the "tipo" of Sipalocyon curtus. However, the specimen referred to in Ameghino's original description of this species is MACN 5960 (-62), which is identified by that name in his catalogue but is not listed as the type. Ameghino (e.g., 1894, fig. 52) figured, as Amphiproviverra manzaniana, a nearly complete palate with complete dentition. This figure is based on the left maxillary and premaxillary with complete dentition of the type of A. manzaniana, MACN 692. There is no trace of the right side nor did Ameghino note its presence in his catalogue. The only part of the upper right dentition preserved with the type is an isolated C. Ameghino apparently restored the right side in his figure by making a mirror image of the left. Fig. 42. Sipaloq/on gracilis Ameghino, 1887 (Santacrucian). PU 15373, left C-M4 of an incomplete skull: a, labial; b, occlusal; c, lingual views. Scale = 30 mm. MARSHALL: REVIEW OF THE HATHLYACYNINAE 59 I -r/ Fig. 43. Sipalocyon gracilis Ameghino, 1887 (Santacrucian). Stereopairs of PU 15373, left C-M4 of an incomplete skull: a, labial; b, occlusal; c, lingual views. Scale = 40 mm. Sipalocyon gracilis is, as noted above, very similar to the Colhuehuapian species S. externa and may be regarded as the direct descendant of that species. The primary changes in this lineage include a slight increase in size, shallowing of the ectoflex, and increase in size of the metacrista on M\ Sipalocyon gracilis is structurally very similar to the contemporaneous species Perathereutes pungens. Both have the P, set at a slight (10°) angle relative to the rest of the tooth row, P, is separated from C and P2 by small but distinct diastems, and talonids on MM are well developed and basined. These species differ in P. pungens being smaller in size, in having relatively larger talonids on MM, and in M, being relatively longer and narrower, and the paraconid being subequal to the talonid in size (latter is larger in S. gracilis). Overall, these differences are minimal, and Perathereutes and Sipalocyon appear to be more closely related than either is to any other contemporaneous genus. These genera shared a common ancestor (possibly in the Deseadan) that was not shared with any other known genus of Hathlyacyninae. 60 FIELDIANA: GEOLOGY Sipalocyon obusta (Ameghino, 1891c). Figures 44-^46; Table 14. Protoproviverra obusta Ameghino, 1891c, p. 313. Amphiproviverra obusta Ameghino, 1894, p. 389. Thylacodictis robusta (sic) Simpson, 1930, p. 51 Type. — MACN 686, a fragment of a left mandibular ramus with base of P3/ M,.2 complete, roots of M3/ M4 complete; and MACN 687, a fragment of a right mandibular ramus (both specimens are of same individual). Hypodigm. — Type only. Horizon and locality. — Collected by C. Ameghino in 1890-91 from the Santa Cruz Formation at Monte Observacion, Santa Cruz Province, Argentina. Age. — Santacrucian. Diagnosis. — Similar in size to S. gracilis, but with a very shallow and gracile mandibular ramus and a very reduced talonid on M4. Comments. — This species is tentatively recognized as valid. In size and struc- ture of the P3-M3 and trigonid portion of the M4, S. obusta is virtually identical to specimens of S. gracilis. However, it differs from specimens of S. gracilis in the M4 having a very reduced talonid and in the very shallow and gracile structure of the mandibular ramus. These features may be a reflection of the young age of this specimen because the posterior molars show no indication of occlusal wear and there are growth lines on the ramus in the area of the M4, suggesting that this element has not yet completed growth. This specimen might thus represent a variant individual of S. gracilis, although it is very unique and there Fig. 44. Sipalocyon obusta (Ameghino, 1891c) (Santacrucian). MACN 686 (type), a fragment of a left mandibular ramus with base of P3, Mw complete, roots of M3, and M4 complete: a, labial; b, occlusal; c, lingual views. Scale = 20 mm. Fie. 45. Sipalocyon obusta (Ameghino, 1891c) (Santacrucian). Stereopairs of MACN 686 (type), a fragment of a left mandibular ramus with base of Pj, M,.2 complete, roots of M3, and M, complete: a, labial; b, occlusal; c, lingual views. Scale = 30 mm. Fig. 46. Sipalocyon obusta (Ameghino, 1891c) (Santacrucian). MACN 686 (type), a fragment of a left mandibular ramus with base of P^ M,.2 complete, roots of Mj, and M, complete: top, occlusal; bottom, lingual views. Scale = 50 mm. 61 3 iriiri 3.2 2 2 * <fc X> 2 <* «N II o 3 I a >** o 3 2 •rt £ *! '•& S °° M Q X in so so so SO t-s 00 00 sO sO 22 UU << 22 62 MARSHALL: REVIEW OF THE HATHLYACYNINAE 63 is nothing else quite like it in the Ameghino collection. Sipalocyon obusta thus apparently represents a species distinct from that of S. gracilis, and they may both have evolved from a Colhuehuapian ancestor like S. externa. Ameghino (1891c, p. 313) orginally placed this species in the Dasyura, family Thylacinidae, but later (1894, p. 389) put it in his group Sparassodonta, family Amphiproviverridae. Notocynus Mercerat, 1891b Notocynus Mercerat, 1891b, p. 81. Type. — Notocynus hermosicus Mercerat, 1891b, p. 81. Distribution. — Montehermosan, Buenos Aires Province, Argentina. Diagnosis. — As for type and only known species. Notocynus hermosicus Mercerat, 1891b. Figures 47, 48; Tables 5, 6. Notocynus hermosicus Mercerat, 1891b, p. 81; Ameghino, 1891f, p. 438; Rovereto, 1914, p. 159; Cabrera, 1927, p. 305, fig. 18; Ringuelet, 1966, p. 55, pi. 11, figs. D, E. Notocynus hermosieus (sic) Mercerat, 1898, p. 58. Didelphys triforata Ameghino, 1891f, p. 438 (partim). Type. — MLP 11-91, a fragment of a left mandibular ramus with roots of C and P,, base of P2, roots of P3-M: (M, crown was lost since the specimen was figured by Cabrera, 1927, fig. 18), M3 present but missing tips of protoconid and para- conid, and anterior root of M4 present. Hypodigm. — Type only. Horizon and locality. — Monte Hermoso, type locality of Montehermosan, Bue- nos Aires Province, Argentina. Age. — Montehermosan. Diagnosis. — Small borhyaenid, similar to Sipalocyon gracilis in size and structure; C weakly developed; tooth row tight, no diastems separating antemolar cheek teeth; PM aligned in same anteroposterior axis; P2 with small but distinct pos- terobasal heel; M,., with large-basined talonid; small anterobasal cingular cusp present on M:3, but not on M,; a large mental foramen opens below anterior root of P2, and another tiny one opens below posterior root of P2. Comments. — Notocynus hermosicus was erected by Mercerat (1891b, p. 81) on a partial left mandibular ramus collected from Monte Hermoso, type locality of the Montehermosan. Ameghino (1891f, p. 438) critically and unjustly attacked Mercerat, noting in his opinion that N. hermosicus was a synonym of Didelphys triforata. The latter, from the same locality as N. hermosicus, was earlier named and figured by Ameghino (1889, p. 280, pi. 22, figs. 37-38). Mercerat later (1898, p. 58) pointed out that Trouessart (1898, p. 1232) accepted N. hermosicus as valid, an opinion contra to that of Ameghino. Ameghino refused, however, to recognize this species, and in his classic work of 1898 Sinopsis geoldgico-paleontoldgica he did not even mention the name N. hermosicus. Cabrera (1927, p. 305, fig. 18) later redescribed and figured the type specimen (MLP 11-91) and rightly concluded that N. hermosicus was indeed valid. Two additional specimens from different horizons and localities than the type have been referred in literature to N. hermosicus. Riggs & Patterson (1939, p. 149) noted that: a portion of maxillary with M! and the mandibular rami of the small borhyaeninae may be referred with confidence to N. hermosicus Mercerat, hitherto known only from the Monte Hermoso beds. . . . 64 FIELDIANA: GEOLOGY Fig. 47. Notocynus hermosicus Mercerat, 1891b (Montehermosan). MLP 11-91 (type), a fragment of a left mandibular ramus with roots of C and Pl7 base of P2, roots of P3-M2, M3 present but missing tips of protoconid and paraconid, and anterior root of M4 present: a, labial; b, occlusal; c, lingual views. Scale = 20 mm. This specimen (FMNH P14409) from beds of Montehermosan age in the province of Catamarca is here (p. 37) named a new species of Borhyaenidium. Reig (1958a, p. 249) tentatively included cf. Notocynus hermosicus in a list of taxa from the Chapadmalal Formation. He later (1958b, p. 280) mentioned that this tentative identification was applied to a cranial fragment (MMP S-240) of a large marsupial that was distinct from other species in the fauna. He pointed out that this specimen was larger than Sparassocynus and did not have an epi- tympanic sinus, but there was evidence of a large alisphenoid bulla. Reig believed this specimen to be of a borhyaenid, and on the basis of its size had earlier (1958a, p. 249) assigned it provisionally to Notocynus. I have not seen this spec- imen, and its true identity is not known. Notocynus hermosicus is very similar to Sipalocyon gracilis (i.e., MACN 5938) from the Santacrucian. In both species there is a small but distinct heel on P2, the M3 has a very large and deeply basined talonid, there is a small anterobasal cingulum, and the size and proportions of the teeth and mandibular ramus are virtually identical. These species differ in N. hermosicus having the P! aligned in the same anteroposterior axis as the rest of the cheek teeth and not being set obliquely as in S. gracilis, and in the P, in N. hermosicus not being separated from MARSHALL: REVIEW OF THE HATHLYACYNINAE 65 Fig. 48. Notoa/nus hermosicus Mercerat, 1891b (Montehermosan). Stereopairs of MLP 11- 91 (type), a fragment of a left mandibular ramus with roots of C and P,, base of P2, roots of Pj-M2, Mj present but missing tips of protoconid and paraconid,and anterior root of M» present: a, labial; b, occlusal; c, lingual views. Scale ■ 30 mm. the C and P2 by diastems. These differences are minor, and I regard S. gracilis as the Santacrucian ancestor of N. hermosicus. Notogale Loomis, 1914 Notogale Loomis, 1914, p. 216. Type. — IPharsophorus mitis Ameghino, 1897, p. 504. Distribution. — Deseadan of Patagonia, and Bolivia. Diagnosis. — As for type and only known species. Notogale mitis (Ameghino, 1897). Figures 49-55; Table 15. IPharsophorus mitis Ameghino, 1897, p. 504. Notogale mitis Loomis, 1914, p. 216, figs. 142, 143; Patterson & Marshall, 1978, figs. 2-7. 66 FIELDIANA: GEOLOGY ■J" Fig. 49. Notogale mitis (Ameghino, 1897) (Deseadan). MACN 52-368 (type), a fragment of a right mandibular ramus with basal portions of M2.3: a, labial; b, occlusal; c, lingual views. Scale = 10 mm. Type. — MACN 52-368, a fragment of a mandibular ramus with basal portions of M2.3 (figured by Patterson & Marshall, 1978, fig. 2). Hypodigm. — The type and AC 3060, a fragment of a left mandibular ramus with P3 and M4 present (figured by Loomis, 1914, fig. 143; Patterson & Marshall, 1978, fig. 3); AC 3117, a left maxillary fragment originally with part of M2, M3 incomplete, and M4 (M4 now missing) (figured by Loomis, 1914, fig. 142; Pat- terson & Marshall, 1978, fig. 6); PU 21867, a fragment of a right mandibular ramus with talonid of M3, M4 complete (figured by Patterson & Marshall, 1978, fig. 4); PU 21868, a fragment of a right mandibular ramus with M3 (missing tip MARSHALL: REVIEW OF THE HATHLYACYNINAE 67 r Fig. 50. Notogale mitis (Ameghino, 1897) (Deseadan). Stereopairs of MACN 52-368 (type), a fragment of a right mandibular ramus with basal portions of \\_ . occlusal view. Scale = 20 mm. of protoconid) (figured by Patterson & Marshall, 1978, fig. 5c, d); PU 21869, a fragment of a right mandibular ramus with roots of C, Pi:, and anterior root of Pv PU 21871, a fragment of a rostrum with roots (or alveoli) of left IM, and base of right C, alveolus of left C, and roots of right P1; PU 21872, a fragment of a left mandibular ramus with P2 (figured by Patterson & Marshall, 1978, fig. 5a, b); PU 21874, a fragment of a left mandibular ramus with base of C and alveoli of I|.v PU 21875, a fragment of a right maxillary with M2 (figured by Patterson & Marshall, 1978, fig. 7, as an M1); PU 21876, a fragment of a left maxillary with M1; PU 21877, greater part of crown of C lacking enamel; PU 21993, a fragment of a right mandibular ramus with bases of M^ and PU 21996, a fragment of a right mandibular ramus with M2 (broken). Horizon and localities. — The type is probably, and the AC specimens are cer- tainly, from Cabeza Blanca, Chubut Province, Argentina. PU 219% is from Bra- nisa Loc. V-5, Salla-Luribay Basin; the other PU specimens are from the general locality of Salla, Bolivia. Age. — Deseadan. Diagnosis. — Similar in size and structure to species of Cladosictis from beds of Colhuehuapian and Santacrucian age of Argentina; upper and lower PI sepa- rated from C and P2 by small but distinct diastemas; PI -3 oriented along same axis as molar series, P) not set obliquely in jaw; P2 longer and narrower than P^ P3 with large, distinct posterobasal cuspule, a feature only faintly developed on Pz; protocone of M1 large, paracone about three-fourths size of metacone; para- cone and metacone approximated, united basally; parastyle small but distinct; stylar shelf virtually absent; protocone of M3 large; shallow ectoflex present; metastylar shear better developed than in M1; lower molars with small but distinct anterobasal cingulum; talonids of Mw distinct, narrower than trigonids; talonids of M3 with distinct hypoconid, entoconid, and hypoconulid; talonid of M< essentially unicusped with enlarged hypoconulid dominant; minute ento- conid and hypoconid visible in unworn teeth (e.g., PU 21867). Comments.— In the type (MACN 52-368) (figs. 49, 50), both teeth have lost all traces of enamel, their bases are fractured, and only a small portion of the mandibular ramus surrounding them is preserved. Their position in the series is not quite certain. Large alveoli can be seen anterior and posterior to them, and their size and relative proportions indicate that they are certainly molars. ■ cs o • ■<* in . t^ oo • tv rv ■> o CS| ~> : : a ^ "* : : : i <s ^ t^ t*+ 2 IH w i— i ID vO H c p t^ rx t-x tv 01 M rH 00 00 00 y E t— t r-l H rH CO (N <N (N M u o Q = O C/l as UJ UDDD i < Cu Oh Cu \0 00 00 OS 00 ON 00 f\|Oi-i<-ii-ii-it-ii-i inrOCNtNrslfNCNfN 2 U <UDDDDDD 68 I I J Fig. 51. Notogale mitis (Ameghino, 1897) (Deseadan). AC 3060, a fragment of a left mandibular ramus with P3 and M4 present: a, lingual; b, occlusal; c, labial views. Scale = 10 mm. 69 70 FIELDIANA: GEOLOGY Fig. 52. Notogale mitis (Ameghino, 1897) (Deseadan). PU 21872, a fragment of a left mandibular ramus with P2: a, occlusal; b, labial views. PU 21868, a fragment of a right mandibular ramus with M3 (missing tip of protoconid): c, occlusal; d, labial views. Scale = 50 mm. It is assumed that the larger tooth is more posterior in the series, and these teeth are identified as M2.3 although they could possibly, but less probably, represent M^. The proportions and size of the roots of the type match the alveoli of AC 3060, figured by Loomis (1914, p. 217, fig. 143). The P2 has been lost since the specimen was figured by him. The two remaining teeth are P3 and IVL,; the latter is clearly not M3 as thought by Loomis, who further complicated matters by labeling it M2. The M4 was probably isolated from the rest of the jaw when the specimen was collected and was restored into the M3 position. Patterson & Marshall (1978, fig. 3) restored M4 in its correct position (fig. 51). In the same box with the type of Notogale mitis (MACN 52-368) is an isolated right P1 of a borhyaenid. This tooth measures 7.3 mm in length and 4.5 mm in width. It is similar in size and structure to specimens of Borhyaena tuberata from the Santa Cruz Formation. Its preservation suggests, however, that it is from Cabeza Blanca and is thus of Deseadan age. Ameghino makes no mention of this tooth, and how it came to be associated with the type of N. mitis is not known. This P1 is clearly not referable to N. mitis, but may prove referable to a species of Pharsophorus, perhaps P. lacerans. Notogale mitis is the most abundant species of Borhyaenidae in the Salla fauna. It is small with a rather generalized dentition, suggesting that it may have been Fie. 53. Notogale milis (Ameghino, 1897) (Deseadan). PU 21867, a fragment of a right mandibular ramus with talonid of Mj, and M« complete: a, lingual; b, occlusal; c, labial views. Scale = 50 mm. Fie. 54. Notogale mitis (Ameghino, 1897) (Deseadan). AC 3117, a left maxillary frag- ment with posterior edge of M2 and M3 very broken: a, labial; b, occlusal; c, lingual views. Scale = 50 mm. 71 72 FIELDIANA: GEOLOGY Fig. 55. Notogale mitis (Ameghino, 1897) (Deseadan). PU 21875, a fragment of a right maxillary with M2: a, labial; b, occlusal; c, lingual views. Scale = 50 mm. a fox-like carnivore. Notogale mitis is similar in size to some large species of Pliocene-Pleistocene Didelphidae, and it is conceivable that the unconfirmed reports of didelphids in the Salla fauna of Bolivia (Hoffstetter, 1968, 1976) and in the type Deseadan fauna at La Flecha (Tournouer, 1903, p. 469) were based on fragmentary remains of this species. Notogale mitis is similar in size and structure to species of Cladosictis from beds of Colhuehuapian and Santacrucian age and may be regarded as ancestral to that genus. Notogale tenuis (Ameghino, 1897) ?Pharsophorus tenuis Ameghino, 1897, p. 504. Notogale tenuis Loomis, 1914, p. 217; Patterson & Marshall, 1978, p. 54 (as nomen vanum). Type. — MACN 52-387, an isolated upper? premolar. Hypodigm. — Type only. Locality. — No specific locality data, but probably from Cabeza Blanca, Chubut Province, Argentina. Age. — Deseadan. Comments. — Ameghino (1897, p. 504) believed this tooth to be a third lower premolar. It is, however, probably an upper, but which one is not certain. The crown is low, and dentine is exposed on the lingual? surface. The tooth measures 3.4 mm long, 2.6 mm wide. Its features suggest possible affinity with other marsupial groups, notably the Caenolestidae and more specifically the Palae- MARSHALL: REVIEW OF THE HATHLYACYNINAE 73 othentinae. In any case, this tooth is clearly not of a borhyaenid, and the taxon is a nomen vanum. Cladosictis Ameghino, 1887 Cladosictis Ameghino, 1887, p. 7; 1889, p. 286. Clasodictis (sic) Roger, 18%, p. 13. Cladictis (sic) Winge, 1923, p. 67. Hathliacynus Ameghino, 1887, p. 7. Agustylus Ameghino, 1887, p. 7. Ictioborus Ameghino, 1891c, p. 315. Type of Cladosictis. — Cladosictis patagonica Ameghino, 1887, p. 7. Type of Hathliacynus. — Hathliacynus lustratus Ameghino, 1887, p. 7. Type of Agustylus. — Agustylus cynoides Ameghino, 1887, p. 7. Type of Ictioborus. — Ictioborus fenest rat us Ameghino, 1891c, p. 315. Diagnosis. — Medium-sized borhyaenids; teeth aligned in a straight line in tooth row, PI not set at angle relative to main horizontal axis of jaw; PI separated from C and from P2 by small but distinct diastems; talonid very reduced and with shallow basin on MM (M4 sometimes only cuspate); P3 large and proodont; trigon on M1"3 with shallow basin in unworn teeth; trigon on M* very reduced and typically cuspate; stylar shelf greatly reduced; small but distinct parastyle present on M'"\ Cladosictis centralis Ameghino, 1902c. Figures 56-59, Table 16. Cladosictis centralis Ameghino, 1902c, p. 128. Type. — MACN 11639, anterior half of an edentulous cranium with roots of right C-M4 and roots of left C-M1. Hypodigm. — Type and MNHN col. 4, a left mandibular ramus with alveolus of C, roots of P, and P2-M4 complete; and a right mandibular ramus with C-M4 complete (only tip of P3 is missing), both of same individual. Horizon and locality. — Both specimens are from the Colhue-Huapi Formation at the Barranca south of Lago Colhu£-Huapi, Chubut Province, Argentina; the type was collected by C. Ameghino and the MNHN specimen by A. Tournouer. Age. — Colhuehuapian. Diagnosis. — Smallest and most generalized species of genus; talonid of M4 distinctly basined. Description. — Rostral region of skull is long and narrow, minimal breadth be- tween P1 2 contact is 18.0 mm; sagittal crest was probably large and well devel- oped, judging from its anteriormost edge, which is preserved; supraorbital pro- cesses are large, are formed by frontals, and measure 21.7 mm between them; nasals are narrow anteriorly, becoming very broad posteriorly in region of nasolacrimal and nasofrontal contacts; nasofrontal contact is broad (breadth ■ 19.5 mm), and nasals form V at this contact; lacrimal bone extends onto rostral region, but lacrimal duct lies completely within orbit; a small rugosity is devel- oped on lacrimals along edge bordering orbit, and distance between these ru- gosities is 31.0 mm; nasolacrimal contact is very broad and measures 7.0 mm on left side and 6.5 mm on right; a large infraorbital canal opens in maxillary just above middle of P3 and has a maximum dorsoventral diameter of 7.0 mm on left side and 7.3 mm on right side; maxillaries are swollen anteriorly by roots of canines; maxillary part of palate is perforated only by small nutrient foramina 74 MARSHALL: REVIEW OF THE HATHLYACYNINAE 75 Fig. 57. Cladosictis centralis Ameghino, 1902c (Colhuehuapian). Stereopairs of MACN 11639 (type), anterior half of an edentulous cranium with roots of right C-M4 and roots of left C-M1: palatal view. Scale = 50 mm. along its entire length, and two well-developed nutrient foramina occur at an- terior edge and just medial to posterior edge of Cs; palatine bones lack nutrient foramina and are solid throughout their length; palatines extend anteriorly to a point between M2s and obtain a combined maximum breadth anteriorly of 11.5 mm and a breadth of 17.5 mm between middle of M3s; C root of right side measures L = 9.0 mm, W = 5.5 mm; breadth of palate between Cs = 10.5 mm. Opposite: Fig. 56. Cladosictis centralis Ameghino, 1902c (Colhuehuapian). MACN 11639 (type), anterior half of an edentulous cranium with roots of right C-M4 and roots of left C-M1: top, dorsal; middle, left lateral; bottom, palatal views. Scale = 50 mm. T <-< co rH J On On 6 ~* CM CM Tj< CO j c\i t-i •, -loin -r m <* -r oo oc t< ts co co »t ** in rx ts rv >C nO CO CO X a^ vC <i c- r-t ^C r^ xC vd 00 ON ni ni Ifi in vC \C CO •* rxi ri in vC \o ^o 01 _; • goo 76 1 J3 i s % £ £ < e So I? •sa 3 IE 8 g a c 77 I I I I I I 78 MARSHALL: REVIEW OF THE HATHLYACYNINAE 79 Premolars increase in size from P1 to P3; P1 is separated from C and P2 by a small diastema (ca. 2.0 mm); posterior root of P3 is very large; some measure- ments of alveoli on right side are: P1 L = 6.4 mm, W = 2.8 mm; P2 L = 7.0 mm, W = 2.9 mm; P L = 8.0 mm, W = 4.7 mm; P*3 L = 23.7 mm; maximum breadth of palate between P's = 10.3 mm; maximum breadth of palate between posterior roots of Ps = 16.3 mm. Molars increase slightly in size from M' to M3; orientation of medial and posterior roots indicates that a significant metastylar shear was developed; M4 is two rooted; some measurements of alveoli on right side include: M' L = 7.7 mm, W = 5.9 mm; M2 L = 8.1 mm, W = 6.3 mm; M3 L = 8.4 mm, W = 6.8 mm; M1 3 L = 23.0 mm; P'-M3 L - 46.0 mm. Comments. — The type of Cladosictis centralis, MACN 11639, is slightly smaller and more gracile than all specimens of C. patagonica with which it was compared. I here regard C. centralis as the slightly smaller but direct ancestor of the San- tacrucian species C. patagonica. In turn, C. centralis probably evolved from the Deseadan species Notogale mitis. Cladosictis patagonica Ameghino, 1887. Figures 60-67; Tables 17-22. Cladosictis patagonica Ameghino, 1887, p. 7; 1889, p. 286; 1894, p. 387; 1905, p. 8, fig. 7; Cabrera, 1927, p. 280, figs. 4, 5. Hathliacynus lustratus Ameghino, 1887, p. 7; 1889, p. 286; 1894, p. 383; 1935, p. 106, fig. 3 (caption); Mercerat, 1891a, p. 53. Cladosictis lustratus Sinclair, 1905, pi. 1; 1906, p. 386, pi. 59, fig. 7; Schlosser, 1925, fig. 55; Piveteau, 1961, fig. 26. Cladosictis lustrata Cabrera, 1927, p. 385, fig. 5. Cladosictis lateralis Ameghino, 1894, p. 388. Proviverra trouessartii Ameghino, 1891a, p. 149, fig. 54; 1891d, p. 354. Cladosictis trouessarti Ameghino, 1894, p. 387, figs. 50, 51; 1898, p. 192, fig. 57c-d; 1904, p. 38, fig. 24; 1906, figs. 189, 190. Protoproviverra ensidens Ameghino, 1891c, p. 313. Amphiproviverra ensidens Ameghino, 1894, p. 389; 1935, p. 107, fig. 5 (caption only). Thylacodictis ensidens Cabrera, 1927, p. 300, fig. 17. Cladosictis petersoni Sinclair, 1906, p. 391, with 14 figures. Hathliacynus fischeri Mercerat, 1891a, p. 52; Ameghino, 1894, p. 383 (as a junior synonym of H. lustratus). Ictioborus fenestrate Ameghino, 1891c, p. 315; 1894, p. 3%, fig. 56; 1898, pp. 191, 193, fig. 58f; 1935, p. Ill, fig. 13 (caption only). Amphithereutes obscurus Ameghino, 1894 fide Roger, 18% {nomen nudum); Ameghino, 1935, p. 108, fig. 8 (caption only). Agustylus minus Ameghino, 1935, p. 109, fig. 22 (caption only) (nomen nudum). Acrocyon sectorius Ameghino, 1889, p. 289n, pi. 1, fig. 19 (partim). Agustylus cynoides Ameghino, 1887, p. 7; 1889, p. 287; 1891c, p. 315; 1894, p. 391, fig. 53; 1898. pp. 191, 193, fig. 58c; 1935, p. 109, fig. 38; Mercerat, 1891a, p. 54; Cabrera, 1927, p. 290, fig. 7. Agustylus primaevus Mercerat, 1891a, p. 54; Cabrera, 1927, p. 291, fig. 8. Hathliacynus cultridens Mercerat, 1891a, p. 53; Cabrera, 1927, p. 280, 291 (as partial synonym of Cladosictis patagonica and Agustylus primaevus, respectively). Hathlyacynus cultridens Ameghino, 1894, p. 387 (as junior synonym of Cladosictis troues- sarti). Opposite: Fig. 59. Cladosictis centralis Ameghino, 1902c (Colhuehuapian). Stereopairs of MNHN col. 4, a right mandibular ramus with C-M4 complete: a, labial; b, occlusal; c, lingual views. Scale = 50 mm. 80 FIELDIANA: GEOLOGY Cladosictis dissimilis Mercerat, 1891a, p. 51; Ameghino, 1891d, p. 354 and 1894, p. 387 (as junior synonym of Proviverra troussartii). Hathliacynus rollieri Mercerat, 1891a, p. 53; Ameghino, 1894, p. 391 (as junior synonym of Agustylus cynoides). Type of Hathliacynus patagonica. — MLP 11-103, a fragment of a left maxillary with M3"4. Type of Hathliacynus fischeri. — MLP 11-19, rostrum of skull and attached man- dible with nearly complete dentition. Type of Cladosictis lateralis. — MACN 5950, a rostral portion of skull with roots of left P-P2, and right P-P3, and with left P3-M4 and right MM complete (this specimen is not listed as "tipo" in Ameghino's catalogue or on the specimen, but the original description fits it perfectly — MACN 5951, the distal end of a humerus, is listed as belonging to the same individual, but Ameghino makes no mention of it in his original description of this species). Lectotype of Hathliacynus lustratus. — MLP 11-17, a right maxillary fragment with M1'2 (originally with a P3, but this has since been lost, see Cabrera, 1927, p. 285). (Mercerat referred to both upper and lower dentitions in the original description of this species. The upper dentition is surely MLP 11-17 and is here recognized as the lectotype. The lower dentition is probably MLP 11-11 because it is listed as cotype of this species in the MLP catalogue, and it agrees perfectly with the original description.) Type of Proviverra trouessartii. — MACN 2079, greater part of left side of skull with partial dentition but missing basicranium, right zygomatic arch, right max- illary, and all of right dentition [listed as type in catalogue and on specimen; figured by Ameghino (1894, figs. 50, 51; 1898, figs. 57c, d; 1904, fig. 24; 1906, figs. 189, 190)]. Fig. 60. Cladosictis patagonica Ameghino, 1887 (Santacrucian). MACN 674, a fragment of a left mandibular ramus with P3-M4: a, labial; b, occlusal; c, lingual views. Scale = 30 mm. Fig. 61. Cladosictis patagonica Ameghino, 1887 (Santacrucian). Stereopairs of MACN 674, a fragment of a left mandibular ramus with P3-M4: a, labial; b, occlusal; c, lingual views. Scale ■ 40 mm. Table 17. Measurements of mandibular rami of Cladosictis patagonica. Depth of ramus Depth of ramus Specimen below labial Breadth of below labial Breadth of side of M, same side of M . same MACN 664 14.5 5.2 MACN 681 14.5 6.0 MACN 6288(1) 20.2 7.7 . . . MACN 6288(r) 21.0 8.3 ... ... MACN 9360 14.0 5.9 14.0 7.2 MACN 9386 . . . 15.0 7.0 MACN 11-2 17.5 7.0 ... ... MACN 11-10 14.5 5.8 ... ... 81 I 1 I I I » Fig. 62. Cladosictis patagonica Ameghino, 1887 (Santacrucian). MACN 664 (type of Ictio- borus fenestrates), a fragment of a right mandibular ramus with posterior alveolus of P1# P2 complete, base of P3, alveoli of Mu M2 complete, and M3 missing protoconid: top, labial; middle, occlusal; bottom, lingual views. Scale = 50 mm. Table 18. Statistics for some dimensions of mandibular rami of Cladosictis patagonica. Dimension N OR X s CV Depth of ramus below labial side of Mj 7 14.0-21.0 16.60 2.98 17.95 Breadth of ramus below M, 7 5.2-8.3 6.56 1.13 17.23 Depth of ramus below labial side of M4 2 14.0-15.0 14.50 0.71 4.90 Breadth of ramus below M4 2 7.0-7.2 7.10 0.14 1.97 82 MARSHALL: REVIEW OF THE HATHLYACYNINAE 83 Fig. 63. Cladosictis patagonica Ameghino, 1887 (Santacrucian). MACN 2079 (type of Pro- xnverra trouessartii), greater part of left side of skull with partial dentition but missing basicranium, right zygomatic arch, right maxillary, and all of right dentition: top, dorsal; middle, left lateral; bottom, palatal views. Scale = 50 mm. Type of Cladosictis petersoni. — PU 15702, rostral part of skull and associated partial skeleton. Type of Protoproviverra ensidens. — MACN 689, a fragment of a right mandibular ramus with posterior half of P, and P2, P.-, complete, and anterior root of M, (listed as "tipo" in Ameghino's catalogue and on specimen). Type of Agustylus minus. — MACN 671, a fragment of a left mandibular ramus with M2_, broken; and MACN 672, a fragment of a right mandibular ramus with roots of M,, M2.j complete, M4 broken; both specimens of same individual (listed as "tipo" both in Ameghino's catalogue and on specimen — the specific name minus is crossed out, and afnoides is written above it). Type of Amphithereutes obscurus. — MACN 681, a fragment of a left mandibular ramus with roots of C-M, [listed as type in Ameghino's catalogue and on spec- 84 MARSHALL: REVIEW OF THE HATHLYACYNINAE 85 imen, and it is only specimen listed by this name in the Ameghino collection. This name is, however, crossed out, and "Sparassodont (Amphiproviverridae)" is substituted above it). Type of Ictioborus fenestratus. — MACN 664, a fragment of a right mandibular ramus with posterior alveolus of P„ P2 complete, base of Pj, alveoli of M„ M: complete, M3 missing protoconid [listed as "tipo" in Ameghino's catalogue and on specimen, and original description fits perfectly — figured by Ameghino (1894, fig. 56; 1898, fig. 58f)). Lectotype of Agustylus cynoides. — MLP 11-78, a fragment of a left mandibular ramus with P2 missing anterior edge and anterior root of P3 (P3 broken since figured by Cabrera, 1927, fig. 7). (Ameghino originally described two specimens, both partial rami with a P3. MLP 11-78 is regarded as the lectotype, and the other specimen, a cotype, is almost certainly MLP 11-100.) Type of Agustylus primaevus. — MLP 11-4, greater part of a right mandibular ramus with C missing tip of crown, P, complete, P2 missing posterior edge, P3- M, complete, alveoli of M2, and Mw complete (figured by Cabrera, 1927, fig. 8B). Lectotype of Hathliacynus cultridens. — MLP 11-2, fragment of a left mandibular ramus with alveoli of I,3, base of C, P,.2 complete, alveoli of P3-M, (made lectotype by Cabrera, 1927, p. 282). Type of Cladosictis dissimilis. — MLP 11-86, a fragment of a right maxillary with alveoli of M1, and M2"* complete. Type of Hathliacynus rollieri. — MLP 11-10, a fragment of a left mandibular ramus with posterior root of P3, and M,_, present (M2_, missing tips of protoconids). Hypodigm. — The 15 types and MACN 85, a left maxillary fragment with P3-M" complete; MACN 86, a right maxillary fragment with posterior half of P3, and M1 J complete; MACN 639, an isolated C; MACN 640, an isolated C; MACN 641, a right maxillary fragment with M1 (latter three specimens are listed as probably belonging to Cladosictis trouessarti); MACN 665, a fragment of a left mandibular ramus with roots of P,3 [listed in Ameghino's catalogue as probably belonging to Ictioborus fenestratus; figured by Ameghino (1889, pi. 1, fig. 19) as Acrocyon sectorius]; MACN 673, fragment of a right mandibular ramus with C-M4 (partially broken); MACN 674, a fragment of a left mandibular ramus with P3-M, (same individual as 673) [this specimen originally had the C and P2 as shown in Ame- ghino (1894, fig. 53; 1898, fig. 58c)); MACN 675a, a fragment of a right man- dibular ramus with MI2, and left M13 (both specimens are of same individual and are listed as Agustylus cynoides in Ameghino's catalogue); MACN 690, a right maxillary fragment with P3 (same individual as 689 and possible cotype of Am- phiproviverra ensidens, but there is no direct mention of this specimen in the original description of that species); MACN 705, a left maxillary fragment with P'-M2 complete, an upper isolated P2?, and two fragments of an upper C; MACN 2080 (possibly same individual as 2079 and possible cotype of Proviverra troues- sartii), a fragment of a left mandibular ramus with roots of M2_,; MACN 5927, nearly complete cranium with teeth (listed as Hathliacynus lustratus in Ame- Opposite: Fie. 64. Cladosictis patagonica Ameghino, 1887 (Santacrucian). MACN 5950 (type of Clad- osictis lateralis), a rostral portion of skull with roots of left I'-P2 and right I'-P\ and with left P3-M4 and right MM complete: top, dorsal; middle, left lateral; bottom, palatal views. Scale = 50 mm. Fig. 65. Cladosictis patagonica Ameghino, 1887 (Santacrucian). MACN 5927, nearly com- plete skull with most of dentition: top, dorsal; middle, left lateral; bottom, palatal views. Scale = 50 mm. 86 MARSHALL: REVIEW OF THE HATHLYACYNINAE 87 ghino's catalogue — it is the most complete skull of this species in the Ameghino collection); MACN 5928, a right mandibular ramus with I, v C, P, complete, posterior root of P2, P3-M2 complete, and roots of M^ MACN 5929, vertebrae and other bones (same individual as MACN 5927 and 5928 — I could not locate MACN 5928 in the Ameghino collection); MACN 5951, distal end of humerus (supposedly same individual as 5950); MACN 6280, a left maxillary fragment with C-M4 complete, and associated postcranial elements (6281-6285)— fragment of cranium with nuchal and sagittal crest (6281), vertebrae (6282, 6283), proximal end of right femur (6284), most of right fibula (6285) — listed as Cladosictis troues- sarti and supposedly of a single individual; MACN 6288-6298 (listed as Anath- erium defossus in Ameghino's catalogue), both mandibular rami with teeth [6288 — figured by Ameghino (1898, fig. 57b). Ameghino noted in his catalogue that this specimen had all the teeth preserved, and in his illustration all are indeed present. Now the specimen is missing the medial portion of the coronoid process on the right side (more than is shown in the illustration), the tip of the P: and all of M2.3 are also missing. The jaw has also been broken below the M4 and M,, respectively. The rest of the dentition is present as figured by Ameghino], a left maxillary fragment with roots of P3, M1'2 complete and alveoli of M3 (6289), isolated teeth (6290-92), and various postcranial fragments (6293-98); MACN 9350, a left maxillary fragment with roots of P\ P2-M* complete, but very worn; MACN 9351, a fragment of a left mandibular ramus with roots of M„ M2.3 complete, but very worn (same individual as 9350; both listed as Cladosictis lateralis in Ameghino's catalogue); MACN 9359, a fragment of a right mandibular ramus with P,.2 complete (listed as Amphiproviverra ensidens in Ameghino's cat- alogue); MACN 9360, a fragment of a right mandibular ramus with M14 complete; MACN 9361, a fragment of a left maxillary with M23; MACN 9386, a fragment of a right mandibular ramus with Mw complete (listed as Agustylus cynoides in Ameghino's catalogue); MACN 9387, a fragment of a right mandibular ramus with M^ MACN 9388, a fragment of a left mandibular ramus with M, (MACN 9387-88 are listed as Agustylus cynoides in Ameghino's catalogue); MLP 11-8, a fragment of a right mandibular ramus with anterior alveoli and posterior half of M„ M2.3 complete, and anterior alveolus of M4 (original cotype of Hathliacynus cultridens); MLP 11-9, a fragment of a left mandibular ramus with M<; MLP 11- 11, a fragment of a right mandibular ramus with alveoli of I,3, base of C, roots of P,2 (probable cotype of C. lustratus); MLP 11-13, greater part of a left man- dibular ramus with P1-M4 complete; MLP 11-15, a fragment of a right mandibular ramus with M3; MLP 11-63, a fragment of a left mandibular ramus with MM complete (labeled Thylacodictis ensidens); MLP 11-68, a fragment of a left man- dibular ramus with root of C, P, complete, roots of P23; MLP 11-87, a fragment of a left maxillary with M2 broken, and M3 complete; MLP 11-100, a fragment of a right mandibular ramus with P2 (probable cotype of Agustylus cynoides; this is the specimen that could not be located by Cabrera, 1927, p. 291); MLP 11-120, a fragment of a left mandibular ramus with alveoli of Pi-M„ and M2 complete; PU 15015 (now UCMP 26617), a right mandibular ramus with C-M<; PU 15046, a skull, mandible, and partial skeleton; PU 15097, a left mandibular ramus with C, alveoli of P,2 (which were erupted, but are missing from specimen), DP3-M2 erupted, and M, erupting (an incompletely developed and unerupted P3 is also preserved and is visible below anterior root of DP3); PU 15170, skull, mandible, and partial skeleton; PU 15704 (now CMNH 2893), a right mandibular ramus o n 't ^< •* o N(N MM N N OifNinM't s 00 00 On t^ . in — rsl . tv • • fS cs <N <N CN • tN co tN • fS . . ts. fN in X in . t^ ~r ON . o\ • • LO ir, in ■>* NO • ■* vC •<* • in . o • NO m C-J co in . i— i a- rt <N rsl • t^ • sC sC nC NO nO • NO sC fx tN NO . 00 o X m ro nO . o m o tS t^ . in • 00 0> X X X oo • oo On 3^ oo oo ■ 00 .oooininKM'j'N •\dininuSinininin Oi— i oo oo o\ csi ^ in OtJOOtNts'tsOONOO : 00 00 .MnnNOOfontN :inininiriinininin** 3 on o in no oo K .oooocsiTjirofNiroON ro ro . o : en . -<i< CO : K K • oo : nO inNOTtONOOONONONONCsirsiromfvNoor^^H ooooNONor^rviininininNONOONON°p0pr7r70N SSSSSSSSSSSSS2SS2S< 88 MARSHALL: REVIEW OF THE HATHLYACYNINAE 89 Table 20. Statistics for some upper cheek tooth dimensions of Cladosictis patagonica. Dimension N OR X s cv P' L 3 4.6-6.0 5.07 0.81 15.98 W 2.0-2.1 2.05 0.07 3.41 P2 L 6.2-6.9 6.50 0.29 4.46 W 2.2-2.7 2.48 0.21 8.47 P3 L 6.8-7.4 7.13 0.27 3.79 W 3.0-3.7 3.31 0.23 6.95 M1 L 14 6.9-8.0 7.36 0.37 5.03 W 13 4.7-5.3 5.12 0.18 3.52 M2 L 13 7.4-8.7 8.12 0.38 4.68 W 12 5.4-6.0 5.70 0.17 2.98 M3 L 12 8.0-9.3 8.67 0.36 4.15 W 11 6.1-7.3 6.62 0.42 6.34 M< L 9 4.7-6.5 5.57 0.66 11.85 W 9 2.2-3.1 2.69 0.26 9.67 P'-M3 L 7 46.3-54.4 48.80 3.77 7.73 M'-3 L 9 22.2-25.0 23.72 0.95 4.01 with PM and MM complete, and M, broken; PU 15556, part of a humerus; PU 15705, fragments of skull and skeleton (collected as float); PU 15831, very frag- mentary skull and postcranial remains; PU 15837, very fragmentary skull and postcranial remains; FMNH 13255, a nearly complete skull with dentition; AMNH 9134, a fragment of a left maxillary with C-M4, a fragment of a left mandibular ramus with P3 broken, M,.2 complete, and Mw broken, partial skull with occipital region, and assorted vertebrae and bone fragments; AMNH 9548, a fragment of a right mandibular ramus with P2-M, complete and M2_, broken, and a fragment of left mandibular ramus with PM and M,.3 complete, and P3 and M< broken; AMNH 9574 (now USNM 5936), a right mandibular ramus with P„ P2 broken, and P3-M3, and a left mandibular ramus with alveoli of C and M2.3, and Pi3, M„ and M4 present. Horizon and locality. — All specimens are from the Santa Cruz Formation, Santa Cruz Province, Argentina, and their localities of collection are as follows: Bar- rancas del rio Santa Cruz MLP 11-78, 11-103 (collected by C. Ameghino, 1887), 11- 103; Santa Cruz MACN 85, 86, 665 (collected by C. Ameghino, 1887), MLP 11-2, 11-8, 11-9, 11-10, 11-11, 11-15, 11-17, 11-120; Karaiken MACN 2079, 2080 (collected by C. Ameghino, 1889-1890); Monte Observacidn MACN 639, 640, 641, 664, 671, 672, 673, 674, 675, 689, 690 (collected by C. Ameghino, 1890-91), MACN 9350, 9351 (collected by C. Ameghino, 1891-92); Rio Gallegos MACN 705 (collected by C. Ameghino, 1890-91); Sehuen MACN 681 (collected by C. Ameghino, 1890-91); La Cueva MACN 9359, 9360, 9361, 9387, 9388 (collected by C. Ameghino, 1892-93); Corriquen-Kaik MACN 5927, 5928, 5950, 5951, 6280-6285 (collected by C. Ameghino, 1892-93); Jack-Harvey MACN 6288-6298, 9386 (collected by C. Ameghino, 1892-93); Monte Leon(?) MLP 11-4, 11-19, 11-86; no locality data MLP 11-13, 11-63, 11-68, 11-87, 11-100; 20 miles south of Coy Inlet PU 15702 (collected by O. A. Peterson, 1899), PU 15015, 15170 (collected by O. A. Peterson, 18%), PU 15046, 15097 (collected by J. B. Hatcher, 1896); Coy Inlet PU 15704 (collected by O. A. Peterson, 1899); Lake Pueyrredon PU 15556, 15705, 15831, 15837 (collected by J. B. Hatcher, 1899); Wreck Flat or Smith's Rock Flat, 10 miles north of Coy Inlet FMNH P13255; Canaddn de las Vacas, 35 miles south of Santa Cruz AMNH 9134, 9548, 9574 (collected by B. Brown, 1899). s > * c •3 90 MARSHALL: REVIEW OF THE HATHLYACYNINAE 91 Fig. 67. Cladosictis patagonica Ameghino, 1887 (Santacrucian). Stereopairs of MACN 6280, a left maxillary fragment with C-M4 complete: a, labial; b, occlusal; c, lingual views. Scale for lingual view = 50 mm. Age. — Santacrucian. Diagnosis. — Slightly larger than Cladosictis centralis; talonid of M4 cuspate. Comments. — The type of Proviverra trouessartii (MACN 2079) consists of the greater part of a left side of a skull with partial dentition, but missing the basicranium, right zygomatic arch, right maxillary, and all of right dentition. This specimen is figured by Ameghino (1894, fig. 50; 1898, fig. 57c; 1904, fig. 24; 1906, fig. 190) in dorsal view in which the left side and posteriormost edge is restored. The canine and incisors were also restored because none are now present on the specimen. The zygomatic arch region is very narrow in the figure, reflecting the fact that the left arch, although almost complete, is crushed in this specimen but in life was certainly much broader. Ameghino restored the skull with the arches crushed, and the specimen and figure are identical in this respect. The palate and dentition of MACN 2079 are figured by Ameghino (1894, fig. 51; 1898, fig. 57d; 1906, fig. 189). Only parts of the left dentition are preserved, and these are precisely illustrated in the figures. Except for the roots of P' 2, all of the right dentition is missing. Ameghino restored the right side by making a mirror image of the left, a point which is readily apparent by comparing teeth on each side, which are broken in exactly the same way. . oo . • ■>* • • o • On • i-H in o o t* ro -^ rji • Os 00 CN CSI O • CO CO t}< tJ< rl< O O O ts 00 ts OOts iriNvooin ts ts t\ 00 K vO (N ^O ID ■* \£> co co co co co co ■* OS CO CO O ** O O O Os in! vri ts! ts! ts." vo O CN ts! ts! .(N OM3\H • co csi csj co . t— I CO CO ' CO CO CO ts. ts, vO 00 \o \d vd \d oo o o vo ts! is! t— i co co -^ in co CO co co co csi <* oo in ^ is! ts! ts! ts! ts! is in is is! . -* m ** • in CSI • CN CSI CN • CS| CO . in in m ,_ in • in in in •in ZZs<Z- OOOOOOi-HOSONDtsOO csi^inincooNcsioooominNooooooo tSlStstSlSOOONCSICSIcOCOCOCOCOCOCOOO \0\0\0\OvOvOin\OvO(>OsONO>OvOvsON 22222Z2222ZZ2Z2hS UUUUUUUUUUUUUUUohCu 92 -J : -i On O : as ? u o : t< : ^ —j : ^ . o ■ oo : ^r in : ^i" : in : "<r . vo m . . ^* : ^ . <* : on 06 : oo '• On : ov : °. ""! . o o ■ oo . rsi : -r -r : ■*»« t : ro : •»* . o qo . oo o . o . in : od K : rx : X : oo : oo . in • ** Is. M • "1 : fi : : ,<s ci en : r*i • M . o <N CM . rr> : t^ : : t< : tvl K : K : *1 ^ fH • ^ : pS rn : (r> '• : H . rx rf . ON : vo vo : v£j • 'I ^ : pS rn . C»> Tt : t< r< • ^ : on On On : fvj (S <n rj . m o o r^ : K K oo vb (snin ri <n <s N(On in in in o o On On" 0> ON in ■<* On o no k o © NO NO r-H Cn| t-H l-< en r<S en ro s tj« o in ro rx t< hNN on oo On O no r< 5 ^■On ■?"- On Cm -5-m •-7 ir\ ** ir. ^ jL ESS. ! ku^D^D 93 94 FIELDIANA: GEOLOGY Table 22. Statistics for some lower cheek tooth dimensions of Cladosictis patagonica. N OR x s CV Dimension Pi L W P2 L W P3 L W M, L W \A L 1VI2 W M L M3 W M L ivi^ W P.-M4 L MM L 12 5.1-5.9 5.38 0.21 3.90 12 2.2-2.6 2.45 0.11 4.49 12 6.7-8.0 7.30 0.37 5.07 13 2.7-3.1 2.89 0.12 4.15 14 7.0-7.8 7.39 0.20 2.71 15 2.8-3.5 3.23 0.18 5.57 16 6.2-7.0 6.68 0.22 3.29 17 2.6-3.3 3.05 0.17 5.57 19 6.4-7.3 6.98 0.21 3.01 19 3.2-3.9 3.50 0.17 4.86 19 7.0-8.5 7.76 0.36 4.64 18 3.5-4.2 3.98 0.17 4.27 14 8.2-9.6 8.94 0.43 4.81 12 4.4-5.1 4.80 0.24 5.00 4 57.0-62.5 59.18 2.61 4.41 9 29.5-33.0 30.79 1.12 3.64 MACN 2080 is listed as "tipo" of C. trouessarti in Ameghino's catalogue and on the specimen. In the catalogue it is indicated that this specimen and MACN 2079 are of the same individual. However, there is no specific reference to MACN 2080, a partial left mandibular ramus, in Ameghino's original description of this species, and it cannot therefore be regarded as an official cotype. Ameghino (1891c, p. 315) diagnosed Ictioborus fenestratus as having two lower incisors, a single-rooted rudimentary Pu P2 being double rooted and very large, P3 somewhat smaller, M,^ similar to species of Sipalocyon "pero con una pequeno callo basal estiliforme sobre la cara anteroexterna del lobulo anterior de cada muela." In the type (MACN 664) as shown in Ameghino (1894, fig. 56; 1898, fig. 58f), the anterior edge of the mandibular ramus, including the anterior root of the Pi and the incisors, is missing. This feature of preservation led Ameghino to believe that the incisors were reduced to two in number (although there is no evidence even for this), and he misinterpreted the posterior alveolus of Pj as a single-rooted Pj. The other features he mentions do not serve to separate this species from any of the other specimens here assigned to Cladosictis pata- gonica. Ameghino (1889, p. 289, pi. 1, fig. 19) figured a fragment of a left mandibular ramus with roots of PM and referred it to Acrocyon sectorius. Mercerat (1891a, p. 55) noted that this specimen did not appear to be referable to either this genus or species. Cabrera (1927, p. 313) later commented that this specimen was ap- parently lost, because he could not locate it in the MLP collection. He also called attention to the fact that this specimen did not appear to be referable to A. sectorius. During the winter of 1974, I found this specimen in the Ameghino collection in the MACN (No. 665) where it was apparently taken by Ameghino. It is clearly referable to Cladosictis patagonica and not to A. sectorius, which is a small Santacrucian member of the subfamily Borhyaeninae (Marshall, 1978, pp. 61-63). Two specimens in the MACN are listed in Ameghino's catalogue as being collected by Carlos Ameghino in 1890-91, but are indicated to be types of species that were erected in 1887. The specimens in question are MACN 673-674, listed as type of Agustylus cynoides, and MACN 705, listed as type of Hathliacynus MARSHALL: REVIEW OF THE HATHLYACYNINAE 95 lustratus. These specimens are clearly not the types that are in the MLR An explanation for this situation exists and is discussed by Marshall (1980, p. 36n). A fragment of a right mandibular ramus (MACN 13205) with a broken but relatively complete M4 was collected from late Tertiary beds along the Rfo Parang near Parana\ Entre Rios Province, Argentina. It was collected from "Parang Entre Rfos, Formaci6n Entrerriana" by Alberto Lelong and acquired by the MACN in 1941. The M4 measures L = 9.0 mm, W ■ 5.0 mm, and it is identical in size and structure to specimens of C. patagonica. The "Formacion Entrerriana" is late Miocene in age (Pascual & Odreman Rivas, 1973, chart opposite p. 318). Pascual & Odreman Rivas (1971, p. 404) note that the predominant faunal elements from these beds indicate a Huayquerian age, although there are also taxa of Santacrucian and Chasicoan age (e.g., Protypo- therium sp., Borhyaena tuberata). These authors interpret the presence of these forms in Huayquerian age beds as being deposited from reworking of beds of Santacrucian and/or Chasicoan age. Such is apparently the case with MACN 13205. Cladosictis patagonica is here regarded as the Santacrucian descendant of the Colhuehuapian species C. centralis and as the Santacrucian ancestor of the Cha- sicoan species Chasicostylus castroi. Chasicostylus Reig, 1957 Chasicostylus Reig, 1957, p. 29. Type. — Chasicostylus castroi Reig, 1957, p. 29. Distribution. — Arroyo Chasic6 Formation, Buenos Aires Province, Argentina. Diagnosis. — As for type and only known species. Chasicostylus castroi Reig, 1957. Figures 68-71; Table 23. Chasicostylus castroi Reig, 1957, p. 29, figs. 1-3; Ringuelet, 1966, p. 55, pi. 11, figs. A-C. Type. — MLP 57-XI-9-2, a fragment of a left maxillary with M1'2, a fragment of right mandibular ramus with posterior half of M, and M23 complete, and an isolated left M,, all of a single associated individual. Hypodigm. — The type and possibly MLP 55-IV-28-59, a fragment of a right mandibular ramus with base of C and alveoli of P13. Horizon and locality. — Both specimens are from the Arroyo Chasicd Formation, Arroyo Chasic6, Buenos Aires Province, Argentina. Age. — Chasicoan. Diagnosis. — Medium-sized borhyaenid, slightly larger than Cladosictis patago- nica; lower molars increase slightly in size from M, to M3, all are very narrow transversely (especially trigonid) and have a small but distinctly basined talonid; small anterobasal cingular cusp present on M2.3, absent on M,; upper molars with small but distinct protocone, paracone fused basally with larger metacone; parastyle very large on M' and extends anteriorly toward P3; parastyle-paracone- metacone on M lie in same axis; parastyle on M shorter and situated labiad of paracone-metacone axis; paracone and metacone lie on labial side of crown on M', more medially on M2; no trace of ectocingulum or ectoflex on M1 or M2; metacrista very well developed. Comments.— Chasicostylus castroi is similar to, and is probably descendant from the Santacrucian species Cladosictis patagonica. In fact, all of the specializations that make C. castroi so unique, as noted by Reig (1957), are incipiently developed Fig. 68. Chasicostylus castroi Reig, 1957 (Chasicoan). MLP 57-XI-9-2 (type), a fragment of a right mandibular ramus with posterior half of Mi and M2.3 complete: a, labial; b, occlusal; c, lingual views. Scale = 20 mm. 96 A J ••'.- Fig. 69. Chasicostylus castroi Reig, 1957 (Chasicoan). Stereopairs of MLP 57-XI-9-2 (type), a fragment of a right mandibular ramus with posterior half of M! and M2.3 complete: a, labial; b, occlusal; c, lingual views. Scale = 20 mm. /«*,▲ Table 23. Measurements of cheek teeth of Chasicostylus castroi. Ml M2 M3 Specimen L W L W L W MLP 57-X1-9-2 Upper dentition Lower dentition(l) Lower dentition(r) 8.6 7.5 4.4 2.7 2.8 8.9 8.2 4.8 3.6 8.4 3.8 97 Fig. 70. Chasicostylus castroi Reig, 1957 (Chasicoan). MLP 57-XI-9-2 (type), a fragment of a left maxillary with M1"2: a, labial; b, occlusal; c, lingual views. Scale = 20 mm. 98 MARSHALL: REVIEW OF THE HATHLYACYNINAE 99 Fig. 71. Chaskostylus castroi Reig, 1957 (Chasicoan). Stereopairs of MLP 57-XI-9-2 (type), a fragment of a left maxillary with M1'2: a, labial; b, occlusal; c, lingual views. Scale = 20 mm. in Cladosictis patagonica, and are simply carried to an extreme in C. castroi. The primary differences between these species include: C. castroi being slightly larger in size, in having the upper and lower molars proportionately more elongated anteroposteriorly, in the upper molars having a larger parastyle that extends more anteriorly, in having the protocone set more anteriad relative to the para- and metacone, and in having a more elongated metacrista that is partially a function of the anterior displacement of the protocone. These differences show C. castroi to have better developed shearing specializations than occur in Cla- dosictis patagonica. Anatherium Ameghino, 1887 Anatherium Ameghino, 1887, p. 8. Acyon Ameghino, 1887, p. 8. Type of Anatherium. — Anatherium defossus Ameghino, 1887, p. 8. Type of Acyon. — Acyon tricuspidatus Ameghino, 1887, p. 8. Distribution. — Colhuehuapian and Santacrucian of Patagonia, southern Ar- gentina. Diagnosis. — Largest known genus of Hathlyacyninae; P, separated from C and from P2 by large diastems: P,3 set in straight line in jaw; talonids of M,.3 very reduced and not basined; MA talonid cuspate; anterobasal cingulum well devel- oped on M2A; posterobasal cuspule weakly developed on P2J. -c •e V, w5 (0 • rv c/5 ^S « > <u — . &, "S bO 3 (8 60 '-' S3 £ ° ^ In (X bO <o c 3 ■» = tn 0> •- 3 £ O U ra a 8 g o E o |"2 •2 m *- ~ 100 « I ■ « I I Fie. 73. Analherium herrerae sp. nov. (Colhuehuapian). Stereopairs of FMNH P13521 (type), greater part of a right mandibular ramus with base of C, and with P1-M4 (tips of Pj.2 and protoconid of M3 are missing): a, labial; b, occlusal; c, lingual views. Scale for labial view = 50 mm. 101 a <H - O I-H cO CO £ s in in ■<* in d d * s i«5 o^ : ^ "*! "^ : 06 CJ\ ON 00 On oq o CO CO ro ■<* O O • °^ 00 00 : od • °) ** "* : tN ro ro •*r on oo od ■ ON rj< ro : c4 ro ro in o On O oo r^ r\i csi in ro Os Ov ro ■* C\i <N 3 > 0 ~ >- oo 3 c CN Cn| in in On 00 T cL ro ro NO ON ^H tn t— i i— i in on no in i-i ^ Pu C 2222 2 XX 'f. U UU U 2 2 t < <<< c 22 PL, LL, T3 2S22 •S "< "< 102 MARSHALL. REVIEW OF THE HATHLYACYNINAE 103 Anatherium herrerae, sp. nov. Figures 72, 73; Tables 24, 25. Cladosictis sp. Sinclair, 1930, p. 39, pi. 8, fig. 4, 4a. Etymology. — herrerae named in honor of Senorita Hebe E. Herrera (Museo de La Plata, La Plata and Universidad de la Patagonia, Comodoro Rivadavia) in recognition of her ongoing studies of the marsupial fauna from, and geology of, the Barranca south of Lago Colhue-Huapi from whence this specimen was col- lected. Type. — FMNH P13521, associated left and right mandibular rami with bases of Cs, and P|-M« (tips of right P, ; and protoconids of right M3 and left Mw missing). Hypodigm. — Type only. Horizon and locality. — Colhue-Huapi Formation at Barranca south of Lago Col- hue-Huapi, Chubut Province, Argentina; collected by C. H. Riggs (son of E. S. Riggs) on Nov. 26, 1923. Age. — Colhuehuapian. Diagnosis. — Largest known species of genus; diastems separating P, from C and P: larger than in Anatherium defossus. Table 25. Measurements of mandibular rami of Anatherium defossus and A. herrerae sp. nov. Specimen Depth of ramus below labial side of M, Breadth of same Depth of ramus below labial side of M . Breadth of same A. defossus MACN9 MACN 5988 MACN 11-64 17.4 20.5 18.7 7.6 8.5 7.3 19.8 8.1 A. herrerae sp. nov. FMNH P13521(l) FMNH P13521(r) 21.0 21.5 9.3 10.0 22.4 10*8 Comments. — Anatherium herrerae is the largest species of Hathlyacyninae known. Apart from its slightly larger size and greater size of the diastems sep- arating P, from C and P2, A. herrerae is identical in dental morphology to the Santacrucian species A. defossus. I therefore recognize A. herrerae as the slightly larger Colhuehuapian ancestor of A. defossus. Anatherium defossus Ameghino, 1887. Figures 74-78; Tables 24, 25. Anatherium defossus Ameghino, 1887, p. 8; 1889, p. 289; 1891c, p. 315; 1891d, p. 354; 1894, p. 384; 1898, p. 192, fig. 57b; 1935, p. 110, fig. 10 (caption only). Anatherium defossus (sic) Palmer, 1904, p. 101. Hathliacynus defossus Mercerat, 1891a, p. 53. Acyon tricuspidatus Ameghino, 1887, p. 8; 1889, p. 290; 1894, p. 397; 1935, p. 110; Roger, 18%, p. 18. Acyon tricuspidactus (sic) Mercerat, 1891a, p. 55. Hathliacynus tricuspidatus Mercerat, 1891a, p. 52. Cladosictis tricuspidata Cabrera, 1927, p. 288, fig. 6. Acyon? bardus Ameghino, 1889, pp. 292, 293, pi. 1, fig. 18. Agustylus bardus Ameghino, 1894, p. 392; 1935, p. 109, fig. 9 (caption only). Ictioborus destructor Ameghino, 1894, p. 3%. 104 FIELDIANA: GEOLOGY Fig. 74. Anatherium defossus Ameghino, 1887 (Santacrucian). MACN 9 (type of Acyon? bardus), a fragment of a left mandibular ramus with roots of M^ M3 complete, M4 missing tip of protoconid (crowns of MM are heavily worn): a, labial; b, occlusal; c, lingual views. Scale = 20 mm. Type of Anatherium defossus. — MACN 669, a fragment of a right mandibular ramus with roots of I,.3, base of C, base of Pu and P2 complete (listed as "tipo" in Ameghino' s catalogue, and original description fits perfectly). Type of Acyon tricuspidatus. — MLP 11-64, a right mandibular ramus in two parts, anterior with root of C, alveoli of P12; and posterior with P3-M4 all present but partially broken (figured by Cabrera, 1927, fig. 6). Type of Acyon? bardus. — MACN 9, a fragment of a left mandibular ramus with roots of M,.2, M3 complete, M4 missing tip of protoconid — crowns of Mw are heavily worn (listed as "tipo" in Ameghino's catalogue and on specimen — fig- ured by Ameghino, 1889, pi. 1, fig. 18). Type of Ictioborus destructor. — MACN 5988, a fragment of a right mandibular ramus with posterior root of Pu roots of P23, M12 complete but very, very worn (this specimen is not listed as type in Ameghino's catalogue, but it is listed as "tipo" on a card with the specimen, and the original description of the species matches this specimen perfectly). Hypodigm. — The four types and MACN 646, a fragment of a left mandibular ramus with posterior half of P3, and roots of M12, base of M3 and anterior alveolus of M4 (listed as Agustylus bardus in Ameghino's catalogue). Fie. 75. Anatherium defossus Ameghino, 1887 (Santacrucian). Stereopairs of MACN 9 (type of Acyonl bardus Ameghino, 1889), a fragment of a left mandibular ramus with roots of M,.2, M3 complete, M4 missing tip of protoconid— crowns of Mj_, are heavily worn: a, labial; b, occlusal; c, lingual views. Scale = 30 mm. 105 106 FIELDIANA: GEOLOGY Fig. 76. Anatherium defossus Ameghino, 1887 (Santacrucian). MACN 9 (type of Acyonl bardus), a fragment of a left mandibular ramus with roots of M,.2, M3 complete, M4 missing tip of protoconid — crowns of MM are heavily worn: top, occlusal; bottom, lingual views. Scale = 30 mm. Horizon and locality. — All specimens are from the Santa Cruz Formation, Santa Cruz Province, Patagonia, Argentina. MLP 11-64 is probably from Monte Leon, MACN 9 is from Santa Cruz (date not specified but probably 1887), MACN 5988 is from Shehuen (1890-91), and MACN 646 and 669 are from Monte Observacion (1890-91); all were collected by C. Ameghino. Age. — Santacrucian . Diagnosis. — Largest known Santacrucian species of Hathlyacyninae; slightly smaller than herrerae; diastems separating P! from C and P2 smaller than in A. herrerae. Comments. — The type of Ictioborus destructor, MACN 5988, is the largest known specimen of Anatherium defossus. The ramus anterior to the posterior root of the Pi is missing, and this feature of preservation led Ameghino to believe that the P, was rudimentary and single rooted. Thus, he placed destructor in the genus Ictioborus, which he characterized as having a single-rooted, rudimentary Pj (see p. 94). MACN 5988 is a very old individual as evidenced by the deeply worn M,.2 in which the pulp cavities are exposed on Mj and nearly so on M2. This specimen has the most worn teeth of any animal in the Ameghino collection, and it may J Fig. 77. Anatherium defossus Ameghino, 1887 (Santacrucian). MACN 5988 (type of Ictio- borus destructor), a fragment of a right mandibular ramus with posterior root of Pu roots of P2.j, and M,.2 complete but very worn: top, occlusal; bottom, labial views. Scale = 50 mm. Fie. 78. Anatherium defossus Ameghino, 1887 (Santacrucian). Stereopairs of MACN 5988 (type of Ictioborus destructor), a fragment of a right mandibular ramus with posterior root of P,, roots of P2.v M|.2 complete but very worn: labial view. Scale = 50 mm. 107 108 FIELDIANA: GEOLOGY represent an aged male. Four mental foramina occur on the labial side of the ramus — a very large one below the P2 and three smaller ones of similar size below the anterior roots of P2, M„ and M2, respectively. This animal has the same pathological disorder as occurs in specimens of Sipalocyon gracilis (see p. 57), as evidenced by the irregular rugosities around the second and third most anterior mental foramina and along the inner surface of the symphysis. Acyonl tricuspidatus was diagnosed by Ameghino (1887, p. 8) as having four tricuspate lower premolars. This, however, is not so — the P3 crown on the type (MLP 11-64) was erroneously restored onto the broken base of the M„ giving the impression that there were four premolars in this specimen. I have restored the P3 into its proper position, and the species tricuspidatus now shows no differences with respect to other specimens here assigned to A. defossus. Hathlyacyninae Indeterminate Simpson (1948, pp. 48-49) recorded the presence of a number of fragmentary specimens and isolated teeth from beds of early Tertiary age in Patagonia. These could not be assigned to named taxa and themselves were inadequate to be made types of new species. Many, if not all, appear to represent generalized borhyaenids, and they may prove referable to the Hathlyacyninae. However, I can add little to Simpson's comments about these specimens, and they must remain in taxonomic limbo. The same is true for the unnamed specimens men- tioned by Paula Couto (1970) from the Riochican fissure fills at Sao Jose de Itaborai, Brazil. Palaeocladosictis mosesi Paula Couto (1961, p. 332, fig. 12) from Itaborai was based on what was called an upper M1 or M2 (MNRJ 2671-V). In the original description it was compared with Cladosictis and Thylacodictis ( = Sipalocyori) and was thus associated with members of the Hathlyacyninae. This tooth is, how- ever, not of a borhyaenid, but represents a DP4 or P4 of some sort of ungulate (Marshall, 1978, p. 70). Summary of Evolution of Hathlyacyninae Members of the borhyaenid subfamily Hathlyacyninae are known from beds of Riochican (late Paleocene) through Montehermosan (early Pliocene) age in Argentina and beds of Deseadan (early Oligocene) age in Bolivia. Eighteen species and 12 genera are recognized: Patene simpsoni, P. coluapiensis, Procladosictis anomala, Pseudonotictis pusillus, Notictis ortizi, Perathereutes pungens, Borhyaenidium musteloides, B. riggsi, Sipalocyon externa, S. gracilis, S. obusta, Notocynus hermosicus, Notogale mitis, Cladosictis centralis, C. patagonica, Chasicostylus castroi, Anatherium herrerae, and A. defossus. Upper dentitions of various species showing relative proportions and size of the teeth are compared in Figure 79, and lower dentitions are compared in Figure 80. A summary of various characters, some diagnostic, for the genera are pre- sented in Table 26. Size distributions as indicated by length of Mw and length versus width of M4 are compared in Figures 81 and 82, respectively. These plots demonstrate that absolute and/or relative size differences are sufficient in most cases to differentiate species of a given age. Based largely on basicranial characters, the Hathlyacyninae have been shown (Marshall, 1978, p. 72, fig. 13) to be a monophyletic group relative to other MARSHALL: REVIEW OF THE HATHLYACYNINAE 109 borhyaenid subfamilies. Dentally, hathlyacynines are the most generalized of known borhyaenids and share a number of features with their presumed di- delphoid ancestors. Included, among other features, are a high incisor number; their generally small to medium size; their long, shallow, and generally gracile mandibular rami; symphysis ligamentous and rami unfused in adults; C weakly or moderately developed; presence of a metaconid in earliest forms (i.e., Patene); M,.j typically with large-basined talonids; and a large protocone and wide stylar shelf on M1'3 in early forms. The general evolutionary trend among hathlyacynines has been the increase in carnassial specializations resulting in loss of metaconid and reduction in size of protocone, stylar shelf, and talonid. A general trend of size increase also occurs, although some recognized lineages show size reduction and others show no significant size change. The general conclusion that can be made regarding the evolution of this group is that it has been extremely conservative with regard to the other borhyaenid subfamilies. Further, if sheer numbers of individuals and taxa are a gauge, then it has also been the most successful. The recognized species of Hathlyacyninae are listed in Figure 83 in order of their chronostratigraphic occurrence and with an indication of their probable phylogenetic relationship. The proposed phylogeny is based largely on the data set in Table 26, although it includes consideration of other features discussed in the text. The character states that occur in Patene are regarded as plesiomorphic for the subfamily. Patene simpsoni is the most generalized of known borhyaenids, and if more ABC ^v^^ ^M£%% t^zs^w *&& px^ c? £? <J9 *& ;_F „ _ <3 0&&% <*&©3> Fig. 79. Comparison of upper dentitions of various species of Hathlyacyninae showing relative size and proportions of teeth. All illustrations are drawn to same scale. A, labial; B, occlusal; C, lingual views. 1, Cladosictis patagonica; 2, Chasicostylus castroi; 3, Notogale mitis; 4, Sipalocyon gracilis; 5, Sipalocyon externa; 6, Borhyaenidium riggsi; 7, Borhyaenidium musteloides; 8, Pseudonotictis pusillus; 9, Procladosictis anomala; 10, Patene coluapiensis; 11, Patene simpsoni. 110 FIELDIANA: GEOLOGY ^ £v3 d^ z> <^\_Jcs2 Q, (SSOQccss^ c^L^cU^ i^SxJ^ ^CO^fecs A_^£^ ^^b^^ C5£53£3£e<s>o«3 ^^^^^ ^^^^i^ (g^B^S^S^ <s> ^S^d^A^ -^-j^^ Cc^S^j^ccX)0 *=* Aftm^-- ^%^^^ OQssssds^csj o° £km£^ Q^CS^L G^£l£C2£S£>p ooo oo *^_j£iv353o53o £*&&& •«eDC3cr>DC»o • ££*23£i ><^££&£> ^^^gj^S^JCS) s> <&&&S&J^ !d& CBCD j <i£i/^Oo /&^P SOaa Q5&3& Fig. 80. Comparison of lower dentitions of various species of Hathlyacyninae showing relative size and proportions of teeth. All illustrations are drawn to same scale. A, labial; B, occlusal; C, lingual views. 1, Anatherium herrerae; 2, Anatherium defossus; 3, Chasicostylus castroi; 4, Cladosictis patagonica; 5, Cladosictis centralis; 6 and 7, Notogale mitis; 8, Notocynus hermosicus; 9, Sipalocycn obusta; 10 and 11, Sipalocyon gracilis; 12, Borhyaenidium riggsi; 13, Borhyaenidium musteloides; 14, Perathereutes pungens; 15, Notictis ortizi; 16, Pseudonotictis pu- sillus; 17 and 18, Patene simpsoni. specialized taxa like cf. Nemolestes sp. were not contemporaneous with it (see Marshall, 1978, p. 27), P. simpsoni would be an ideal prototype for the superfamily Borhyaenoidea. Nevertheless, P. simpsoni is structurally the most generalized and temporally the oldest of known Hathlyacyninae and may thus be regarded as representing the basal stock for the subfamily. Patene simpsoni is the only known species of Hathlyacyninae of Riochican age. B E E sc H Ifffll J 4 6 li § ill III it;* = c s Si ■11 i n 2 : : i i _ a .£ _ »> .5 i -V 2 i -a 2 - f £ .5 f E S. li i Is 1 II I * 1 1 6 22 II 2 3 v I* I V ! 22 i 6 _ ^ HI f ir~i = c " s §i =5 8. I 1 -c2 f 2 2 .5 . £ f | | .11 ll| .1 i. c pa li! - «r « if XI 2 '3 ^a*| rnilr III 1 . £ 1 li II -off 1153 h Li22 J ,,1 .-1 M £1 J 1 js-S \ 111 CO O CO CM 00 CM - c J ^3- 1^ CM CO CM CM CM CO CM CM CM o>_ 00_ r^_ •I T3 ^3 O «, * SX 3 • in >> CO •— CO CO •-~ .w TO .CO co TO CO CD > ^ ^ .N c CD CO O) ;6 c TO CO -c o OQ to -Q O c o O o TO Q. CO •^ .O 3 TO . . . t/i .2 -H- 3 c 1 5 i— i >^ 00 * (j « a | CO Q. .CO O o c d> CO o. c o .CO o o 0> c Q. CO to "5 Q) v. TO v. a> CL 5 5 c TO CO i«s o o CO k. D) c o o o to Q. CO 1 TO TO O) O O =2 TO k. C CD O .CO O CO O ■o TO O C o O) TO TO Q. .co O CO O ■o TO O co CO O •*— TO ■o E "C CD TO C CD -C °c CD TO C CQ •~" 112 MARSHALL: REVIEW OF THE HATHLYACYNINAE 113 The Casamayoran species Patene coluapiensis differs from P. simpsoni in being about one-fourth to one-third larger in linear tooth dimensions, in having rel- atively smaller protocones and a slightly narrower stylar shelf on M1 \ a more reduced metacone on M\ and a slightly deeper ectoflex on M\ These taxa are very similar in other respects, and considering their occurrence in successive- aged beds, I recognize P. coluapiensis as the slightly larger and more specialized Casamayoran descendant of P. simpsoni. Procladosictis anomala from beds of Mustersan age has a relatively smaller and cuspate protocone than in species of Patene, and the M3 has a very deep ectoflex. The latter feature is not found in any other borhyaenid and excludes P. anomala as a potential ancestral form for any known later taxa. Procladosictis anomala is here regarded as a slightly larger, specialized, dead-end descendant of the Casa- mayoran species Patene coluapiensis. A metaconid is definitely known only in Patene simpsoni and by inference in P. coluapiensis as well. An isolated lower molar possibly referable to the Mustersan genus Procladosictis also has a metaconid (see p. 18), whereas all known post- Mustersan hathlyacynines lack a metaconid. Loss of this structure thus occurred early in the evolutionary history of this group, although the exact time of loss cannot yet be firmly established. It is, however, assumed in the phylogeny proposed in Figure 83 that metaconid loss occurred independently at least two times within the Hathlyacyninae. Pseudonotictis pusillus is the smallest species of borhyaenid known, being slightly smaller in most comparable linear tooth dimensions than the Riochican species Patene simpsoni (compare measurements in tables 1 and 2). Pseudonotictis pusillus differs from Patene simpsoni in having more advanced carnassial special- izations (i.e., more reduced protocone, stylar shelf, and talonid; loss of meta- conid; elongation of M,; and increase in size and importance of postvallum- prevallid shear). These changes are nevertheless minor considering the time separating these species, and I choose to regard Patene simpsoni as the probable structural ancestor of P. pusillus. Notictis ortizi is the smallest species of post-Santacrucian borhyaenid known and is very similar to Pseudonotictis pusillus. Both species are of similar size with virtually identical molar structure, with weakly developed canines, and with mental foramina of similar size, shape, and position. In N. ortizi, the premolars are absolutely and relatively smaller than in P. pusillus, and the premolar region is foreshortened, the premolars crowded, and the P, is set obliquely in the jaw. The differences in N. ortizi represent but slight specializations of features seen in P. pusillus, and an ancestral-descendant relationship for these species seems probable. Perathereutes pungens from the Santacrucian is intermediate in size between the contemporaneous smaller Pseudonotictis pusillus and the larger Sipalocyon gracilis. I propose that Perathereutes pungens shared a common ancestry with Pseudonotictis pusillus and that this ancestor was characterized by somewhat reduced talonid basins, lack of a metaconid, and presence of small but distinct diastems between C, P„ and P2. Perathereutes pungens is, in turn, here regarded as the most likely known ances- tor for the Huayquerian species Borhyaenidium musteloides. Both species have P, set at a slight oblique angle in the jaw and separated from the C and P2 by diastems, and the talonids (especially on M«) are comparably more reduced relative to other similar-sized Hathlyacyninae. Borhyaenidium musteloides differs §..» F O ^: 3 O O TO 2 o ... Q. O) c oo U) in / <D m •c ^ "QA o CO T3 ^ -d £ T3 c -C O ►X £ C ra jj T3 s C fo !i X -C E c G — o» e <?, o 3 c '— o >, s £ •n VI u x 7J C c Id « 3 C 3 •3 3 SL en — * 01 §"? « H £S *« C o « •y j- c re §05 o -^ fl MN 3 0 cr 9 C w .9 c S'S. in £ 01 J5 .a o tn (j . 60 o c '5 114 MARSHALL: REVIEW OF THE HATHLYACYNINAE 115 South American Land Mammal Aga Montana rmoaan Borhyaemdium nggs, Notocynus harmosicus Huayqoanan Notictia ortizi Borhyaemdium musleloides Chaaicoan Fnaaian Chastcostytus castroi Santacrucian Pseudonotictis Perathereutes Sipalocyon Sipalocyon pusillus pungens obusta gracilis CtadOSiCtiS Anathenum patagonica delossus Cladosictis Anatherium centralis herrerae Fig. 83. Dendrogram showing probable phylogenetic relationships of the genera and species of Hathlyacyninae. from P. pungens in being slightly larger in size and in the talonid basin being proportionately and absolutely more reduced. The Montehermosan species Borhyaenidium riggsi differs from B. musteloides in having, in most cases, slightly larger linear tooth dimensions, a slightly larger M« talonid, and a slightly larger paracone. In other respects these species are inseparable, and they are regarded as representing a single evolutionary lineage. It is proposed that Perathereutes and Sipalocyon shared a common ancestor characterized by a P, set at a slight angle relative to the rest of the tooth row, small diastems separating C, P„ and P2, and talonids on MM being well devel- oped and basined. The Colhuehuapian species Sipalocyon externa is slightly smaller and more gracile than the Santacrucian species S. gracilis, but morphologically these species are virtually indistinguishable. Consequently, I recognize S. externa as the prob- 116 FIELDIANA: GEOLOGY able Colhuehuapian ancestor of S. gracilis. The primary changes in this lineage include slight increase in size, shallowing of the ectoflex, and increase in size of metacrista on M3. Sipalocyon gracilis is the most abundant borhyaenid in the Santa Cruz fauna of Patagonia, southern Argentina. A contemporaneous species, S. obusta, is known only from its type and differs from the large comparative sample of S. gracilis in having a very shallow and gracile mandibular ramus and a very reduced talonid on M4. In other respects S. obusta is indistinguishable from S. gracilis. Sipalocyon obusta is tentatively recognized as a valid species, although it may represent a variant individual of S. gracilis. Both S. gracilis and S. obusta can be easily derived from the Colhuehuapian species S. externa. The Montehermosan species Notocynus hermosicus is very similar to Sipalocyon gracilis; both species are of comparable size, they have a small but distinct heel on P2, M3 has a very large and deeply basined talonid and a small anterobasal cingulum, and the size and proportions of the teeth and mandibular rami are virtually identical. These species differ in N. hermosicus having P, aligned in the same anteroposterior axis as the rest of the tooth row and not being set obliquely as in S. gracilis and in the Pj in N. hermosicus not being separated from the C and P2 by diastems. These differences are minor, and I regard S. gracilis as the probable Santa crucian ancestor of N. hermosicus. The remaining Hathlyacyninae are recognized as monophyletic and can ul- timately be derived from the Casamayoran species Patene coluapiensis. These taxa of the genera Notogale, Cladosictis, Chasicostylus, and Anatherium are of medium to large size, P2 ~ P3, metaconid is absent, talonid of M4 is very small with a shallow basin or cuspate, stylar shelf is very reduced or absent, and talon on M1"4 is moderately developed with a shallow basin or cuspate. The Deseadan species Notogale mitis is a similar size to Patene coluapiensis and can be derived from the latter with a slight reduction in size of the protocone and stylar shelf. Cladosictis and Anatherium shared a post-Deseadan common ancestor that was of large size, had C, Pv and P2 separated by large diastems, and had more reduced protocones and talonid basins than N. mitis. Cladosictis centralis from the Colhuehuapian is known only by a few specimens. These specimens are smaller than the large comparative sample of the Santa- crucian C. patagonica. In other respects these species are inseparable, and I recognize C. centralis as the slightly smaller Colhuehuapian ancestor of C. pa- tagonica. The Chasicoan species Chasicostylus castroi is here regarded as a direct de- scendant of the Santacrucian species Cladosictis patagonica. These species differ in C. castroi being slightly larger in size, in having the upper and lower molars proportionately more elongated anteroposteriorly, in the upper molars having a larger parastyle that extends more anteriorly, in having the protocone set more anteriad relative to the para- and metacone, and in having a more elongated metacrista. These differences show C. castroi to have better developed shearing specializations than occur in Cladosictis patagonica. The species of Anatherium are the largest members of Hathlyacyninae known, and the talonids of M,^ are more reduced than in species of Cladosictis. The Colhuehuapian species A. herrerae is the largest member of the genus, and the diastems separating P, from C and P2 are larger than in the Santacrucian species A. defossus. Both A. herrerae and A. defossus are poorly known, and their specific separation will require reconsideration pending discovery of larger sample sizes. MARSHALL: REVIEW OF THE HATHLYACYNINAE 117 For the present, A. herrerae is regarded as the slightly larger Colhuehuapian ancestor of A. defossus. LITERATURE CITED Amechino, F. 1887. Enumeration sistematica de las espeties de mamiferos fosiles colec- cionados por Carlos Ameghino en los terrenos eocenos de la Patagonia austral y de- positados en el Museo La Plata. Bol. Mus. de La Plata, 1: 1-26. . 1889. Contribuci6n al conocimiento de los mamiferos fdsiles de la Republica Ar- gentina, obra escrita bajo los auspicios de la Academia National de Cientias de la Republica Argentina para presentarla a la Exposition Universal de Paris de 1889. Actas Acad. Cienc. Cbrdoba, VI, xxxiii+ 1,027 pp; atlas, 98 pis. -. 1891a. Caracteres diagnostics de tincuenta espeties nuevas de mamiferos f6siles argentinos. Rev. Arg. Hist. Nat., 1: 129-167. 1891b. Adiriones al eximen de los mamiferos f6siles, tratados en al articulo IV anterior por German Burmeister. Rev. Arg. Hist. Nat., 1: 259-270. -. 1891c. Nuevos restos de mamiferos fbsiles descubiertos por Carlos Ameghino en el Eoceno inferior de la Patagonia austral. Espeties nuevas, aditiones y correctiones. Rev. Arg. Hist. Nat., 1: 289-328. 1891d. Observariones criticas sobre los mamiferos eocenos de la Patagonia austral. Rev. Arg. Hist. Nat., 1: 328-380. -. 1891e. Los monos f6siles del Eoceno de la Republica Argentina. Rev. Arg. Hist. Nat., 1: 383-397. . 1891f. Sobre la supuesta presentia de creodontes en el Mioceno superior de Monte- Hermoso. Rev. Arg. Hist. Nat., 1: 437-438. 1894. Enumeration synoptique des espeties de mammiferes fossiles des formations eocenes de Patagonie. Bol. Acad. Cienc. C6rdoba, 13: 259-452. 1897. Mammiferes Cretaces de 1' Argentine. (Deuxieme contribution a la connais- sance de la faune mammalogique des couches a Pyrotherium.) Bol. Inst. Geog. Argentine 18: 406-521. 1898. Sinopsis geol6gico-paleontologica. Segundo censo de la Republica Argentina. Fol., Buenos Aires, 1: 112-255. 1902a. L'age des formations sedimentaires de Patagonie. An. Soc. Cien. Argentina; L: 109-130; 145-165; 209-229 (1900). LI: 20-39; 65-91 (1900). LII: 189-197; 244-250 (1901). LIV: pp. 161-180; 220-249; 283-342 (1902). 1902b. Notice preliminaires sur les mammiferes nouveaux des terrains cretaces de Patagonie. Bol. Acad. Cienc. C6rdoba, 17: 5-70 (1902-04). -. 1902c. Premiere contribution a la connaissance de la fauna mammalogique des couches a Colpodon. Bol. Acad. Cienc. C6rdoba, 17: 71-141 (1902-04). 1904. Paleontologia argentina. Publicariones de la Universidad de La Plata. Fa- culdad de cientias h'sico-matimaticas, no. 2: 1-79. . 1905. La faceta articular inferior unica del astragalo de algunos mamiferos, no es un caracter primitivo. An. Mus. Nac. Buenos Aires, (3)5: 1-64. . 1906. Les formations sedimentaires du cretace superieur et du Tertiare de Patagonie avec un parallele entre leurs faunes mammalogiques et celles de L'antien continent. An. Mus. Nac. Buenos Aires, (3)8; 1-568. -. 1935. Los Esparasodontes ... A. J. Torcelli (ed.), Ameghino Obras, 19: 103-140 (published posthumously). Burmeister, G. 1891. Aditiones al examen critico de los mamiferos fosiles tratados en el articulo IV anterior. An. Mus. Nac. Buenos Aires, 3: 375-400. Cabrera, A. 1927. Datos para el conocimiento de los dasyuriodeos f6siles argentinos. Revta. Mus. La Plata, 30: 271-315. Hoffstetter, R. 1968. Un gisement de mammiferes Deseadiens (Oligocene inferieur) en Bolivie. C R. Acad. Sti. Paris, 2670: 1095-1097. . 1976. Rongeurs caviomorphes de 1'Oligocene de Bolivie. I. Introduction du De- seadien de Bolivie. Palaeovertebrata, Montpellier, (3)7: 1-14. 118 FIELDIANA: GEOLOGY Kirsch, J. A. W. 1977. The comparative serology of Marsupialia. Austral. J. Zool., Suppl. Sen, 52: 1-152. Kraglievich, L. 1934. La antiguedad pliocena de las faunas de Monte Hermoso y Cha- padmalal, deducidas de su comparacion con las que le precedieron y sucedieron. Mon- tevideo; El Siglo Illustrado, no. 938: 1-136. Lemoine, V. 1891. Etude d'ensemble sur les dents des mammiferes fossiles des environs de Reims. Bull. Soc. Geol. France, ser. 3, 19: 263-290. Loomis, R B. 1914. The Deseadan Formation of Patagonia. Concord, New Haven, pp. 1-232. Marshall, L. G. 1976a. Notes on the deciduous dentition of the Borhyaenidae (Marsu- pialia: Borhyaenoidea). J. Mammal., (4)57: 751-754. . 1976b. Evolution of the family Thylacosmilidae, fossil marsupial "sabertooths" of South America. PaleoBios, 23: 1-30. 1976c. Fossil localities for Santacrucian (Early Miocene) mammals, Santa Cruz Province, Southern Patagonia, Argentina. J. Paleont., (6)50: 1129-1142. -. 1978. Evolution of the Borhyaenidae, extinct South American predaceous mar- supials. Univ. Calif. Publ. Geol. Sri., 117: 1-89. 1979. Review of the Prothylacyninae, an extinct subfamily of South American "dog-like" marsupials. Fieldiana: Geology, n.s. 3: 1-50. -. 1980. Systematics of the South American marsupial family Caenolestidae. Field- iana: Geology, n.s. 5: 1-145. Marshall, L. G., R. F. Butler, R. E. Drake, G. H. Curtis and R. H. Tedford. 1979. Calibration of the Great American Interchange. Science, 204: 272-279. Marshall, L. G., W. A. Clemens, R. J. Hoffstetter, R. Pascual, B. Patterson, R. H. Tedford and W. D. Turnbull. 1977. Acyonidae Ameghino, 1889 (Mammalia): proposed suppression under the Plenary Powers. Bull. Zool. Nomen., (3 + 4)33: 212-213. . 1978. Acyonidae Ameghino, 1889 (Mammalia): supplement to proposal to suppress this name. Z.N.(S.) 2159. Bull. Zool. Nomen., (1)35: 12-14. Marshall, L. G., R. Hoffstetter and R. Pascual. In Press. Geochronology of the continental mammal-bearing Tertiary of South America, Chap. 7. In M. O. Woodburne (ed.), Cenozoic Mammals; Their Temporal Record, Biostratigraphy, and Biochronol- ogy. Univ. Calif. Press. Marshall, L. G., R. Pascual, G. H. Curtis and R. Drake. 1977. South American geo- chronology: radiometric time scale for Middle to Late Tertiary mammal-bearing horizons, Patagonia. Science, 195: 1325-1328. Mercerat, A. 1891a. Caracteres diagnosticos de algunas especies de Creodonta conser- vados en el Museo de La Plata. Rev. Mus. La Plata, 2: 51-56. . 1891b. Sobre un maxilar inferior de Creodonta de Monte Hermoso. Rev. Mus. La Plata, 2: 80-81. 1898. Sur de nouveau restes fossiles de carnassiers primitifs de Monte Hermoso. An. Soc. Cient. Argent., 47: 56-60. Palmer, T. S. 1904. Index Generum Mammalium: a list of the genera and families of mammals. U.S. Dept. Agr., North American Fauna, no. 23: 1-984. Pascual, R. and A. R. Bocchino. 1963. Un nuevo Borhyaenidae (Marsupialia) del Pliocene medio de Hidalgo (La Pampa). Ameghiniana, (4)3: 97-107. Pascual, R. and O. E. Odreman Rivas. 1971. Evolurion de las comunidades de los vertebrados del Terriario argentine Los aspectos paleozoogeograficos y paleoclimaticos relarionados. Ameghiniana, 8: 372-412. . 1973. Las unidades estratigraficas del Terriario portadoras de mamiferos. Su dis- triburion y sus relariones con los acontecimientos diastroficos. Actas del Quinto Con- greso Geol6gico Argentino, 3: 293-338. Patterson, B. 1965. The auditory region of the borhyaenid marsupial Cladosictis. Breviora, 217: 1-9. Patterson, B. and L. G. Marshall. 1978. The Deseadan, Early Oligocene, Marsupialia of South America. Fieldiana: Geology, (2)41: 37-100. MARSHALL: REVIEW OF THE HATHLYACYNINAE 119 Paula Couto, C. de. 1952a. Fossil mammals from the beginning of the Cenozoic in Brazil. Marsupialia: Polydolopidae and Borhyaenidae. Amer. Mus. Novit., 1559: 1-27. . 1952b. Fossil mammals from the beginning of the Cenozoic in Brazil. Marsupialia: Didelphidae. Amer. Mus. Novit., 1567: 1-26. -. 1961. Marsupiais fbsseis do Paleoceno do Brazil. An. Acad, brasil. Cienc., 33: 321-333. 1962. Didelfideos fosiles del Paleoceno de Brasil. Rev. Mus. Argent. Cienc. Nat. "Bernardino Rivadavia/' Cienc. Zool., (112)8: 135-166. -. 1970. News on the fossil marsupials from the Riochican of Brazil. An. Acad, brasil. Cienc. (1)42: 19-34. PrvETEAU, J. 1961. Marsupialia. In J. Piveteau (ed.), Traite de Paleontology. (1)6: 585-637. Reig, O. A. 1952. Observaciones sobre Notictis ortizi Ameghino. Anal. Soc. Gent. Arg., (1)CUV: 3-9. . 1957. Nota previa sobre los marsupiales de la formaci6n Chasic6. Ameghiniana, (3)1: 27-31. 1958a. Notas para una actualization del conocimiento de la fauna de la formaci6n Chapadmalal. I. Lista faunfstica preliminar. Acta. Geol. Lilloana, 2: 241-253. -. 1958b. Notas para una actualizaridn del conocimiento de la fauna de la formaci6n Chapadmalal. II. Amphibia, Reptilia, Aves, Mammalia (Marsupialia: Didelphidae, Bor- hyaenidae). Acta Geol. Lilloana, 2: 255-283. Rices, E. S. and B. Patterson. 1939. Stratigraphy of the late Miocene and Pliocene deposits of the province of Catamarca (Argentina) with notes on the faunae. Physis, 14: 143-162. Ringuelet, A. B. de. 1953. Revisi6n de los didelfidos f6siles argentinos. Revta. Mus. Eva Peron (La Plata] (n.s.), 3: 265-308. . 1966. Marsupialia. In A. V. Borrello (ed.), Paleontografia Bonaerense. Com. Invest. Ci. Prov. Buenos Aires, 4: 46-59. Roger, O. 18%. Verzeichnis der bisher bekannten fossilen Saugethiere. Neu zusammen- gestellt von Dr. Otto Roger. (I. Theil). Ber. Naturw. Ver. Schwaben u. Neuburg, 32: 1-40. Rovereto, C. 1914. Los estratos araucanos y sus fbsiles. An. Mus. Nac. Hist. Nat., Buenos Aires, 25: 1-250. Schlosser, M. 1923. Grundzuge de Palaeontologie (Palaeozoologie) von Karl A. von Zittel. II. Abteilung-Vertebrata. Neuarbeitet von F. Broili und M. Schlosser. Munich and Berlin, R. Oldenbourg, v + 706 pp. . 1925. Class V. Mammalia, Vol. III. Textbook of Paleontology, by K. A. von Zittel, revised by Max Schlosser. MacMillan and Co., London, 316 pp. Scott, W. B. 1937. A History of Land Mammals in the Western Hemisphere. Revised edition, rewritten throughout. New York, The MacMillan Co., xiv + 786 pp. Simpson, G. G. 1930. Post Mesozoic Marsupialia. In Fossilium Catalogus. 1: Animalia. Berlin, W. Junk, Pt. 47: 1-87. . 1935. Descriptions of the oldest known South American mammals, from the Rio Chico Formation. Amer. Mus. Novit., 793: 1-25. . 1945. The principles of classification and a classification of mammals. Bull. Amer. Mus. Nat. Hist., 85: 1-350. . 1948. The Beginning of the Age of Mammals in South America. Part 1. Bull. Amer. Mus. Nat. Hist., (1)91: 1-232. . 1967. The Ameghinos' localities for Early Cenozoic mammals in Patagonia. Bull. Mus. Comp. Zool., (4)136: 63-76. Sinclair, W. J. 1905. The marsupial fauna of the Santa Cruz beds. Proc. Amer. Phil. Soc, XLIV: 73-81. 1906. Mammalia of the Santa Cruz beds: Marsupialia. Rept. Princeton Univ. Exped. Patagonia, (3)4: 333-460. 1930. New carnivorous marsupials from the Deseado Formation of Patagonia. Field Mus. Nat. Hist., Geol. Mem., (1)1: 35-39. 120 FIELDIANA: GEOLOGY Tournouer, A. 1903. Notes sur la geologie et la paleontologie de la Patagonie. Bull. Soc. Geol. France, ser. 4, 3: 463-473. Trouessart, E. L. 1898. Catalogus Mammalium tarn viventium quam fossilium. Neu. ed. Berlin, pt. 5: 665-1264. . 1904. Catalogus Mammalium tarn viventium quam fossilium. Quinquinale sup- plementum, Berlin, Vol. I, pp. 1-546 (1904); Vol. II, pp. 547-929 (1905). Winge, H. 1923. Pattedyr — Slaegter. 1, Monotremata, Marsupialia, Insectivora, Chiroptera, Edentata. Copenhagen, viii + 360 pp., lllpls. [English Translation: The interrelation- ships of the mammalian genera. Copenhagen, 3 vols. (1941-42). (Transl. by E. Deich- mann and G. M. Allen.)]. Field Museum of Natural History Roosevelt Road at Lake Shore Drive Chicago, Illinois 60605 Telephone: (312) 922-9410