CO en Ut.iVLrtSUY OF ^h ILLIKOIS LIBRARY a- AI URBANA-CHAMPAIGN oo GEOLOGY The person charging this material is re- sponsible for its return to the library from which it was withdrawn on or before the Latest Date stamped below. Theft, mutilation, and underlining of books are reasons for disciplinary action and may result in dismissal from the University. To renew call Telephone Center, 333-8400 UNIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN L161— O-1096 FIELDIANA V Geology Published by Field Museum of Natural History New Series, No. 10 SYSTEMATICS OF THE SOUTH AMERICAN MARSUPIAL FAMILY MICROBIOTHERIIDAE LARRY G. MARSHALL THE LIBRARY OF THE JUN 9 1983 UNIVERSITY OP ILi March 31, 1982 AT URBANA-CHAMPAIGN Publication 1331 SYSTEMATICS OF THE SOUTH AMERICAN MARSUPIAL FAMILY MICROBIOTHERIIDAE FIELDIANA Geology Published by Field Museum of Natural History New Series, No. 10 SYSTEMATICS OF THE SOUTH AMERICAN MARSUPIAL FAMILY MICROBIOTHERIIDAE LARRY G. MARSHALL Assistant Curator of Fossil Mammals Department of Geology Field Museum of Natural History Accepted for publication July 10, 1981 March 31, 1982 Publication 1331 Library of Congress Catalog Card No.: 81-69756 ISSN 0096-2651 PRINTED IN THE UNITED STATES OF AMERICA CONTENTS List of Illustrations vi List of Tables vii Abstract 1 Introduction 1 Techniques of Study 2 Abbreviations 2 Acknowledgments 3 Historical Review 4 Systematics 9 Superfamily Didelphoidea 9 Microbiotheriidae 9 Microbiotherium Ameghino, 1887 10 M. acicula (Ameghino, 1891) 12 M. patagonicum Ameghino, 1887 14 M. praecursor (Ameghino, 1898) 19 M. division (Ameghino, 1902e) 22 M. tehuelchum Ameghino, 1887 26 M. gallegosense Sinclair, 1906 35 Dromiciops Thomas, 1894 35 D. australis (Philippi, 1893a) 35 Summary of Evolution of Microbiotheriidae 46 Review of Supposed Microbiotheres 54 Superfamily Indeterminate 54 Ideodelphys Ameghino, 1902d 54 /. microscopicus Ameghino, 1902d 54 Superfamily Didelphoidea 55 Coona Simpson, 1938 55 C. pattersoni Simpson, 1938 55 Eomicrobiotherium nov. gen 57 E. gaudryi (Simpson, 1964) 58 E. gutierrezi (del Corro, 1977) 60 Pachybiotherium Ameghino, 1902e 61 P. acclinum Ameghino, 1902e 61 Superfamily Caenolestoidea 64 Stilotherium Ameghino, 1887 64 S. dissimile Ameghino, 1887 64 Phonocdromus Ameghino, 1894 69 P. gracilis Ameghino, 1894 69 Literature Cited 71 LIST OF ILLUSTRATIONS 1. Map of southern tip of South America showing vertebrate fossil localities dis- cussed in text 3 2. Microbiotherium acicula (Ameghino, 1891) (Santacrucian) 13 3. Microbiotherium patagonicum Ameghino, 1887 (Santacrucian) 15 4. Microbiotherium patagonicum Ameghino, 1887 (Santacrucian) 16 5. Microbiotherium praecursor (Ameghino, 1898) (age unknown) 20 6. Microbiotherium divisum (Ameghino, 1902e) (Colhuehuapian) 23 7. Microbiotherium divisum (Ameghino, 1902e) (Colhuehuapian) 24 8. Microbiotherium tehuelchum Ameghino, 1887 (Santacrucian) 27 9. Microbiotherium tehuelchum Ameghino, 1887 (Santacrucian) 28 10. Microbiotherium tehuelchum Ameghino, 1887 (Santacrucian) 30 11. Microbiotherium gallegosense Sinclair, 1906 (Santacrucian) 36 12. Map of south central Chile and western Argentina showing known geographic range of Dromiciops australis 38 13. Dromiciops australis (Philippi, 1893a) (Recent) 39 14. Dromiciops australis (Philippi, 1893a) (Recent) 42 15. Dromiciops australis (Philippi, 1893a) (Recent) 43 16. Dromiciops australis (Philippi, 1893a) (Recent) 44 17. Dromiciops australis (Philippi, 1893a) (Recent) 45 18. Comparison of upper dentitions of various species of Microbiotheriidae showing relative size and proportions of teeth 47 19. Comparison of lower dentitions of various species of Microbiotheriidae showing relative size and proportions of teeth 47 20. Size distribution of the species of Microbiotherium as indicated by length of Mj_4 48 21. Size distribution of the species of Microbiotherium as indicated by relationship of various linear tooth dimensions 49 22. Dendrogram showing probable phylogenetic relationships of the genera and spe- cies of Microbiotheriidae 53 23. Ideodelphys microscopicus Ameghino, 1902d (Casamayoran) 55 24. Patene pattersoni Simpson, 1938 (Casamayoran) 56 25. Eomicrobiotherium [nov. gen.] gaudryi (Simpson, 1964) (Casamayoran) 58 26. Eomicrobiotherium [nov. gen.] gaudryi (Simpson, 1964) (Casamayoran) 59 27. Eomicrobiotherium [nov. gen.] gutierrezi (del Corro, 1977) (Casamayoran) 60 28. Pachybiotherium acclinum Ameghino, 1902e (Colhuehuapian) 62 29. Stilotherium dissimile Ameghino, 1887 (Santacrucian) 65 30. Stilotherium dissimile Ameghino, 1887 (Santacrucian) 66 31. Stilotherium dissimile Ameghino, 1887 (Santacrucian) 68 32. Phonocdromus gracilis Ameghino, 1894 (Santacrucian) 70 33. Phonocdromus gracilis Ameghino, 1894 (Santacrucian) 71 LIST OF TABLES 1. Classification of Microbiotheriidae as proposed by various workers 5 2. Measurements of lower cheek teeth of Microbiotherium acicula 13 3. Measurements of cheek teeth of Microbiotherium patagonicum 16 4. Measurements of lower cheek teeth of Microbiotherium divisum 23 5. Measurements of upper cheek teeth of Microbiotherium tehuelchum 29 6. Measurements of lower cheek teeth of Microbiotherium tehuelchum 29 7. Statistics for some cheek tooth dimensions of Microbiotherium tehuelchum 31 8. Measurements of cheek teeth of Dromiciops australis 40 9. Statistics for some cheek tooth dimensions of Dromiciops australis 40 10. Summary of some diagnostic characters of known species of Microbiotheriidae 50 11. Measurements of lower cheek teeth of Eomicrobiotherium gaudryi 59 12. Measurements of lower cheek teeth of Pachybiotherium acclinum 63 13. Measurements of cheek teeth of Stilotherium dissimile 67 14. Statistics for some lower cheek tooth dimensions of Stilotherium dissimile 67 15. Measurements of lower cheek teeth of Phonocdromus gracilis 70 ABSTRACT Members of the South American marsupial family Microbiotheriidae are known from Late Oligocene (Colhuehuapian) and Early Miocene (Santacrucian) age beds in Patagonia, southern Argentina, and in the Recent fauna from Chilo6 Island, south central Chile, and from adjacent parts of Argentina. Two genera, Microbiotherium and Dromiciops, and seven species are recognized. Microbio- therium divisum (Ameghino, 1902e) is known only from the Colhuehuapian. M. acicula (Ameghino, 1891); M. patagonicum Ameghino, 1887; M. tehuelchum Ame- ghino, 1887; and M. gallegosense Sinclair, 1906 are known only from the Santa- crucian. M. praecursor (Ameghino, 1898) is of unknown age and provenance, while Dromiciops australis (Philippi, 1893a) is known only from the Recent. These genera and species are distinguished primarily on the basis of absolute size, although relative size and minor structural differences in the canines, cheek teeth, and auditory bullae also have proven diagnostic. The taxonomy of other didelphoid or supposed didelphoid taxa either referred to the Microbiotheriidae or implicated in the ancestry of this family are discussed. These include: Ideodelphys microscopicus Ameghino, 1902d (Casamayoran-Early Eocene); Coona pattersoni Simpson, 1938 (Casamayoran): Eomicrobiotherium [nov. gen.] gaudryi (Simpson, 1964) (Casamayoran); E. gutierrezi (del Corro, 1977) (Casamayoran); and Pachybiotherium acclinum Ameghino, 1902e (Colhuehuapian). It is demonstrated that microbiotheres were among the syntypes of the caeno- lestid species Stilotherium dissimile Ameghino, 1887 (Santacrucian) and Phonoc- dromus gracilis Ameghino, 1894 (Santacrucian). INTRODUCTION A major adaptive radiation of the marsupial superfamily Didelphoidea has been and continues to be documented for the South American Cenozoic. Aspects of this group's evolutionary history have been clarified in recent years, although many problems regarding the phylogenetic relationships of recognized taxa still remain. Existing problems are compounded by the fact that many of the taxa erected by the eminent Argentine paleontologist Florentino Ameghino some 80 to 90 years ago have not been adequately restudied since that time. The sys- tematics of these taxa have yet to be stabilized at the generic and specific levels. Until this is done, the evolutionary history of South American didelphoids in particular and marsupials in general will remain more speculative than is really necessary. One of the didelphoid groups most in need of such a review is the Micro- biotheriidae. Nearly a score of species grouped in a dozen genera have been 2 FIELDIANA: GEOLOGY referred by various workers to this family; many have been regarded in recent years as nomina vana or nomina dubia. This paper presents a detailed systematic revision of living and fossil members of the South American marsupial family Microbiotheriidae. It reviews the tax- onomic history of the Microbiotheriidae, discusses the possible phylogenetic relationships of the included taxa, stabilizes the group's taxonomy at the generic and specific levels, and represents an attempt to bring together in one place an updated and expanded systematic treatment of these animals. The systematics and phylogenetic positions of other didelphoid and caenolestoid taxa which 1) were formally assigned at one time or another to the Microbiotheriidae; 2) were implicated in the ancestry of the Microbiotheriidae; or 3) had specimens of microbiotheres included erroneously as syntypes in original descriptions, also are discussed and the systematic problems clarified. For all of these taxa, an attempt is made to give complete listings of generic and specific synonymies, complete literature citations, and hypodigms inclusive of all known specimens. During the course of this study I was able to examine firsthand nearly all pertinent materials, including type and referred specimens. This work includes discussion and description of some new materials, but essentially is based on a reappraisal of previously known specimens and literature, and gives revised diagnoses for the family, genera, and species. TECHNIQUES OF STUDY Microbiotheres, as in living Didelphidae, have an adult dental formula of Ii-2~i-!-o/ Q, Pi-5-3/ M{:§:|:|. The highest number of teeth known for any fossil or living marsupial, this formula is regarded as primitive (plesiomorphic) for Marsupialia (e.g., Marshall, 1979, p. 388). This conventional system for serial designation of the teeth is used in this study; it is intended to be descriptive and does not necessarily imply homology. The P3 of this system has a prede- cessor or "milk tooth" here referred to as dP3. The Argentine fossil localities mentioned below are shown in Figure 1. The chronology and usage of South American land mammal ages follows Marshall et al. (in press). Specimens were measured with a pair of dial calipers to the nearest 0.1 mm when possible. All measurements are in millimeters (mm). 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: AMNH, American Museum of Natural History, New York; MACN, Museo Argentino de Ciencias Naturales "Bernardino Rivadavia", Buenos Aires; MLP, Museo de La Plata, La Plata, Argentina; MMP, Museo Municipal de Ciencias Naturales de Mar del Plata "Lorenzo Scaglia", Mar del Plata, Argentina; MNHN, Museum National d'Histoire Naturelle, Paris; and PU, Princeton University, Princeton, New Jersey. MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE Fig. 1. Map of southern tip of South America showing vertebrate fossil localities (circles) discussed in text. ACKNOWLEDGMENTS I am indebted to the following individuals for allowing me to study specimens at their respective institutions: R. H. Tedford and M. C. McKenna (AMNH); G. del Corro (MACN); R. Pascual (MLP); G. Scaglia (MMP); R. Hoffstetter (MNHN); and D. Baird (PU). For helpful comments on various phases of this study and/or for reading the manuscript I thank J. A. W. Kirsch, M. C. McKenna, R. Pascual, O. A. Reig, 4 FIELDIANA: GEOLOGY R. H. Tedford, and W. D. Turnbull. The figures of dentitions were drawn by Elizabeth Liebman from original specimens or from epoxy casts. This study's initial stages were made possible by three grants (1329, 1698, 1943) from the National Geographic Society, Washington, D.C., and its comple- tion was made possible by National Science Foundation grant DEB-7901976. HISTORICAL REVIEW The family Microbiotheriidae1 was created by Ameghino (1887, p. 6) and was diagnosed as: Marsupiales sumamente pequenos con un gran incisivo en la mandibula inferior y siete muelas en serie continua, que aumentan de tamano desde la primera a la quinta y disminuyen desde la sexta a la septima.2 Ameghino placed this family in the order Marsupialia and included within it two newly recognized genera and three species, all erected on the basis of partial lower dentitions: Microbiotherium patagonicum, M. tehuelchum, and Stilotherium dissimile (table 1). The major features uniting these taxa were their small size, supposed presence of a large procumbent incisor, overall similarity in structure of the molars, similarity in relative size of the molars to each other, and a small M4. This same taxonomic arrangement was followed by Ameghino, who in 1889 (p. 263) (table 1) further included the family in a new order Microbiotheria diagnosed as: No mas de 5 incisivos. Caninos j. Muelas no menos de f a | con tuberculos puntia- gudos. Paladar con cavidades palatinas. Un interparietal. Arco cigomatico delgado. Craneo puntiagudo. Una rama externa del canal alveolar de la mandibula inferior. Cola larga. Sacro de dos vertebras. Los cuatro miembros iguales o casi iguales. Un hueso intermedio del carpo. Dedos segundo y tercero del pie posterior, no reunidos por un tegumento externo. Plantigrados. Huesos marsupiales presentes. Solo existe en nuestro suelo como representante de este orden, la familia de los Microbiotheridae . It would appear from the diagnosis that relatively complete cranial and post- cranial material were then known. If this were the case, however, then this material was subsequently never described nor figured by Ameghino. In 1891, Ameghino removed Stilotherium from the Microbiotheriidae and re- defined the family as: Incisivos en numero variable, un canino, tres premolares y cuatro molares, todos en serie continua, sin discontinuidad; incisivos pequenos; canino de tamano variable y a menudo premolariforme; el tamano de los tres premolares, pequenos y parecidos a los de Didelphys, aumenta del primero al ultimo. Verdaderos molares inferiores parecidos a los de Didelphys, con dos conos externos y tres internos, pero cuyo tamano disminuye constantemente del primero al ultimo. M4 con el talon posterior simple, no bipartido arriba como en Didelphys. Ramas mandibulares de borde inferior casi recto aumentando gradualmente el alto de adelante hacia atras. 1 Ameghino spelled the family-group name Microbiotheridae. This is a lapsus; according to Article 29 of the International Code of Zoological Nomenclature (see Stoll et al., 1961, 1964), the correct spelling is Microbiotheriidae. This emendation was made by many early workers and by Simpson (1929, p. 116) when he proposed the subfamily rank Microbio- theriinae. 2 Ameghino numbered the premolars and molars as one series and called them all molars. « 5 e 5 .3 Q 3 £ u « & £ S « 5 B « a. a. -5 £ E E £ E EtS •3 -2 -3 -3 -3 -3 a ^J V V QJ -s .« -s: -s: C C « *< -2 -2 -2 -2 -2 -2 o -a -S .S -S _5 2 2 g S 5* S. o o .2 •5 * -2 ~ E .C -^ « o rH Q. O c JU a "3 if) 3 « 60 <u e •o 3 :C 0» _2 n "3 C y :E a S e -C E 0 .2 1£ o 5 H U u •3 .3 s « •2 S t -5= B « E S..2 S"g E -c -3 -s: .2 8 2 2 £ a. ^-^ , — * , — ..«; <j w ui m vj 33_3CP hEhE-KE-KE-K.2 U) U) (A [Q IA >> >, 3^ IS, >> * * -C -C •£ a.a.6.a.a. ^ ^ ^ ^ ^ •3 -3 -S "3 "S o o o o o uj uj uj uj uj ' ■ _ — u •rl a.. 2 "3 -a •S 2 "B •- C-t?C-};C*;C-ijC O f4> O ^> 2 .2 2 .2 .<-> .Q .VJ .£} 5 2^2 e .3 ■§« ss s.s -3?fc .5 £ .0 e .0 s a g ■* 2 S2 ^ -a -S- £§;3.g;3.g Si"! -•2 ^--2 S--2 w--2 ii--2 ^2--2 "-1 3 o £ - H 8 B < S a 3 ^ » w 2 "73 "o '2 3 5 S .gu 3 .£ E •£ 3 g--.c 3S5h <« w «» .y S ^ ,0 3 3 3 c .3 ^^* -k * fc c « w u u u w c a. a.^; j; ^ a.. 2 rsrs 2 2 2^2 uj uj a: a: a: s < 5 a E 3 e 5 a s 5 a 1 e 5 "5 Bo "5 a .^ s S i 5 a ■j-j E 1 S "3 .'0 y g 2 E 5 S a. a a 3 a a S S. 3^ ^ C G I | <A <A <n >> 3^ >, -s ■;S ^ «e -C -= •c ~3 ~: ."5 .6 0 . - 6 ^-2-2-^3 s IS -a 0 >> •2P-5 ■a p s a 0 6 5 ^ o<S k. >^ ». •. a, <x d, a. Cr> S-8 1 E 3 E fc u a 5 3 -P 3 5 -2 u 60 *» i: « 3 CO CO 1-1 rj * -2 -2 O a O -a -a c ,2S££u ^ a. o .y .y r~ < S vi tq 73 _ 3 3 3 J5 J5 JC C — u o vj Sj * •« >! >. 3^-g a. a.^; ^: j; .2 "51 "i! fc J? t -^ :s:s 2 2 2 g ■w Ti 13 t! t! .U uj uj a: a. a: -^ E E •3 -3 5^ .5 3 a .a. « o a. "^s W V) vi 2 3-> >, 3i 3 •«•«•«•£ -5--S--S- 3 ^3TS?3 OO •« T3 •« ^ g g g £> o, a, a, to 6 FIELDIANA: GEOLOGY He further noted that microbiotheres share features with both the Diprotodonta (Garzonidae = Caenolestidae) and the Pedimana (Didelphyidae). As such he re- garded microbiotheres as the basal stock ("tronco antecesor") of the diprotodont and polyprotodont marsupials. Within the Microbiotheriidae he included several new taxa (table 1): Stylognathus diprotodontoides, Microbiotherium forticulum, Eodi- delphys fortis, E. fatnula, Prodidelphys acicula, P. pavita, P. obtusa, Hadrorhynchus tortor, H. torvus, and H. conspicuus. This arrangement remained relatively unchanged for the next seven years. In 1898 Ameghino synoptically rediagnosed the group and the included genera as follows: MICROBIOTHERIDAE. Parecidos a los Didelphyidae, pero con la boveda del paladar sin prolongarse atras de la ultima muela; cuarta a sexta muelas inferiores con dos tuberculos externos y tres internos, ultima inferior mas pequena y canino a menudo poco diferenciado de los incisivos. Aparecen en el Cretaceo inferior y adquieren su mayor desarrollo en el Eoceno superior, donde se extinguen. Proteodidelphys praecursor (n. gen., n. sp.) con algunos caracteres de ungulado, incisivos de corona ancha y excavada como en ciertos ungulados; canino vertical, pequeno y de raiz deprimida perpendicularmente en el centro; las tres primeras muelas inferiores simples sobre el lado externo y complicadas en el interno; tamano muy pequeno; Cretaceo inferior (areniscas abigarradas) de Patagonia. Microbiotherium Ameghino, con el tuberculo an- terior interno de las muelas inferiores atrofiado, tres especies: Microbiotherium pata- gonicum, tehuelchum y forticulum Ameghino, y restos indeterminados en la formacion Guaranitica. Stylognathus diprotodontoides Ameghino, mandibula muy gracil y con un solo incisivo a cada lado, formando una transicion a los Diprotodontes. Eodidelphys Ameghino, se distingue por el tuberculo anterior interno bien desarrollado y el canino muy pequeno, vertical y apenas diferenciado de los incisivos y de la primera muela, dos especies: Eodidelphys fortis y famula Ameghino. El genero Prodidelphys Ameghino, se distingue por el canino recto, vertical, muy pequeno, con un callo basal posterior y apenas un poco mas elevado que el diente que le sigue; la primera, sequnda y tercera muelas estan formadas por un simple cono con un callo basal posterior, tres especies: Prodidelphys acicula, pavita y obtusa Ameghino, las tres de tamano sumamente pequeno. Hadrorhynchus Ameghino, se distingue por la sinfisis muy fuerte, muy des- cendente hacia abajo y prolongada hacia adelante, canino muy robusto: Hadrorhynchus tortor, torvus y conspicuus Ameghino (Ameghino, 1898, p. 187). By 1904, faunas containing marsupials of Casamayoran (Early Eocene) and Col- huehuapian (Late Oligocene) age were known in Patagonia. From the former, Ameghino (1902d, p. 43) described Ideodelphys microscopicus; from the latter, Oligobiotherium divisum (1902e, p. 124), Pachybiotherium acclinum (1902e, p. 123), and Clenia minuscula (1904c, p. 260). These four taxa he referred to the Micro- biotheriidae. At about that time other workers began either to study microbiotheres or at least to reach alternative conclusions based on Ameghino' s published works. Bensley (1903, p. 208) suggested, for example, that It is extremely probable that at least some of the Miocene Microbiotheriidae of South America were in reality Didelphidae, and were the ancestors of the Epanorthidae [ = Caenolestidae] and their allies and of the Sparassodonta [ = Borhyaenidae]. Sinclair (1906) likewise concluded: Although placed by Ameghino in a separate family, the Microbiotheridae, this genus [Microbiotherium] possesses so many important characters in common with the Di- delphyidae that the propriety of its reference to the latter family seems beyond ques- tion (p. 408). Among living forms it approaches most closely some of the subspecies of Caluromys laniger (p. 412). MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 7 Winge (1923 [1941]) placed Microbiotherium in a new tribe Microbiotheriini in the family Epanorthidae ( = Caenolestidae; see Marshall, 1980). He characterized the genus and tribe as having a typical didelphid dental formula of J:}:^:|. However, largely because of the reduction of the stylar shelf and cusps, relative enlarge- ment of the protocone, reduction in size of the paraconid, and small size of the M4, he regarded Microbiotherium as a basal stock for the Epanorthidae and ac- cordingly included it within that family. This classification was not followed by subsequent authors. Simpson (1929, p. 116) regarded the microbiotheres as merely an aberrant group of didelphids, and formally recognized them as a subfamily, Microbio- theriinae, of Didelphidae. In 1935 (p. 136) he diagnosed the subfamily on the basis of dental characters: Premolars trenchant. Metacone somewhat larger than paracone (except in M4). No distinct metastylar spur. External cingulum moderate, of nearly equal width through- out, styles weak. M4 reduced and with narrow, nearly unibasined, unicuspid heel unlike that of other subfamilies. In his classification of 1945, Simpson (pp. 41-42) recognized the microbiotheres as a subfamily, Microbiotheriinae, of Didelphidae and included within it Ideo- delphys and Microbiotherium (which included as junior synonyms Proteodidelphys and Pachybiotherium). He noted (p. 42, n. 1), however, that six or seven of Ame- ghino's other taxa may also belong here, although they are poorly characterized and were regarded by Simpson as nomina vana. Simpson clearly stated that until these taxa were revised their taxonomic position would remain uncertain. Ringuelet (1953) was the first person since Ameghino to attempt a revision of fossil didelphoids of Argentina. She recognized a subfamily Microbiotheriinae and included within it the following taxa: Clenia minuscula, Eodidelphys acicula, E. famula, E. fortis, E. obtusa, E. pavita, E. praecursor, Microbiotherium (Microbio- therium) acclinum, M. (M.) gallegosense, M. (M.) patagonicum, M. (M.) tehuelchum, M. {Hadrorhynchus) conspicuus, M. (H.) tortor, M. (H.) torvus, Microbiotheridion hernandezi, and Ideodelphys microscopicus. She created a new genus, Microbio- theridion, for the Colhuehuapian species hernandezi which Simpson (1932b, p. 2) earlier referred to Microbiotherium. Her recognition of two subgenera of Micro- biotherium (Microbiotherium and Hadrorhynchus) was also a first in this paper. At the time of Ringuelet's study it was believed that microbiotheres were an extinct group known only from fossil deposits in Argentina. It thus came as a surprise when O. A. Reig (1955a) convincingly demonstrated that the living opossum Dromiciops australis from south central Chile and adjacent parts of Argentina was a member of this group. Based on comparisons with the original description and illustrations of Microbiotherium tehuelchum by Sinclair (1906, p. 410, pi. 62, fig. 7), Reig placed Dromiciops in the Microbiotheriinae. He pointed out the similarity in the size and structure of the auditory bullae of these taxa and of the similarity in swelling of the periotic ( = entotympanic of Segall, 1969b) and its participation in formation of the ossified bulla posteriorly. Reig also noted the similarity in development of the elongated premaxillaries and in the form and orientation of the incisors. He further gave detailed comparative descriptions of both upper and lower dentitions. He added that species of the living opossum genera Caluromys, Caluromysiops, and Glironia agree more closely in dental struc- ture and in some particularities of the skull with Dromiciops and Microbiotherium 8 FIELDIANA: GEOLOGY than with other living opossums. Accordingly, Reig regarded these genera,3 albeit somewhat tentatively, as specialized microbiotheres and included them in the Microbiotheriinae. In the same year, Reig (1955b) presented a brief but seminal synthetic review of fossil members of the marsupial family Caenolestidae. He demonstrated that many of these taxa were erected on specimens including microbiotheres. Re- alization of this fact presented another obfuscation to an already miasmatic problem with regard to taxonomic diversity and stability of microbiotheres. This confusion was the result of similarities in lower molar morphology of fossil caenolestids here referred to the species Stilotherium dissimile and Pho- nocdromus gracilis and the microbiothere species Microbiotherium patagonicum and M. tehuelchum (see below). These similarities were responsible for Ameghino's (1887) initial referral of Stilotherium dissimile to the Microbiotheriidae, although he soon realized this error and in 1891 removed Stilotherium to the family Gar- zonidae ( = Caenolestidae) where it clearly belongs (Marshall, 1980). A detailed comparison of the dentition of Stilotherium dissimile with microbiotheres is pre- sented by Pascual & Herrera (1975), a study made with consideration of the didelphid-caenolestid dichotomy. Segall (1969a,b) made detailed comparisons of the auditory region of Dromi- ciops with that of Microbiotherium and with a large number of living didelphids. He convincingly demonstrated the uniqueness of Dromiciops among living forms as well as its remarkable similarity with Microbiotherium. He also demonstrated that the auditory regions of the supposed microbiotheres Caluromysiops, Calu- romys, and Glironia are very similar to each other and to Didelphis, and are notably different from those of Dromiciops and Microbiotherium. Segall (1969b, p. 497) concluded that ... in so far as the auditory region is concerned the similarities [of Caluromysiops, Caluromys and Glironia] with Didelphis are so close that I cannot agree with Reig (1955[a]) who considered them specialized Microbiotherinae (sic) rather than Didelphinae. Furthermore, the tympanic region of Caluromys, while distinct from that of Dro- miciops, shows little relationship to that of Didelphis (Reig & Simpson, 1972). The chromosomes of Dromiciops australis have been studied by Hayman & Martin (1974, p. 7), Sharman (1973, p. 492), and Spotorno & Fernandez (1971, p. 40), and have been shown to have the primitive (plesiomorphic) marsupial number of 2n = 14. Dromiciops alone of all didelphids studied has 22 autosomal arms, and all but one pair of its autosomes are metacentric or submetacentric. Chromosomes five and six are submetacentric as in Caluromys (Sharman, 1973, p. 492, fig. 4). The karyotypes of Glironia and Caluromysiops are unknown. As the karyotypes of species of Marmosa are very similar to those of Caluromys and Dromiciops (Hayman & Martin, 1974), cytology throws little conclusive light on the affinities of these forms. Unfortunately, there is no comparative information yet available on the struc- ture of the reproductive organs or on the morphology and behavior of sperm in Dromiciops. These characters clearly separate Caluromys from most other di- 3 As a further historical note it should be mentioned that Thomas (1894) originally suggested that, of the living opossums, Dromiciops is probably most closely related to Caluromys. In support of such a relationship Osgood (1943, p. 49) noted that, in both, the lining of the pouch is deep cinnamon rufous in color in sharp contrast to the surrounding parts. MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 9 delphids (Hill & Fraser, 1925; Biggers, 1966; Biggers & DeLamater, 1965; and Biggers et al., 1965). Kirsch (1977) has recently used serological evidence to demonstrate that Ca- luromys is not a microbiothere, but instead represents a distinct line of didelphid evolution. He thus erected for it a new subfamily, Caluromyinae. Other evidence for separating Caluromys from the didelphines is their unique pairing of sper- matozoa (Biggers & DeLamater, 1965) and structure of the urogenital tract (Hill & Fraser, 1925; Biggers, 1966). Structure of these features is unknown in Calu- romysiops and Glironia. The evidence from serology (Kirsch, 1977) is that Dromiciops is very distinct from all other didelphids. In addition, Reig (in press; and personal communi- cation as cited in Kirsch & Calaby, 1977) has subsequently and independently concluded that the dental similarities between Dromiciops and the other putative microbiotheres (e.g., Glironia, Caluromys, Caluromysiops) are only convergently similar to Dromiciops and Microbiotherium, though undoubtedly distinct from typical didelphines. This view is more fully documented by Reig, Kirsch, and Marshall (in preparation). In this study I follow recent workers and recognize a family Microbiotheriidae which includes only Dromiciops australis among the living opossums. Of the fossil taxa I recognize only one genus, Microbiotherium, and six species. SYSTEMATICS Order MARSUPIALIA Illiger, 1811, p. 75 Superfamily DIDELPHOIDEA (Gray, 1821, p. 308) Osborn, 1910, p. 515 Family MICROBIOTHERIIDAE Ameghino, 1887, p. 6 ( = Microbiotheria Ameghino, 1889, p. 263; Microbiotheriinae Simpson, 1929, p. 116) (Including Clenialitidae Ameghino, 1909, p. 204) Diagnosis. — Very small to medium sized opossums; If, C\, Pi M\; mandibular rami unfused along symphysis which extends posteriorly to position below C-P] contact; large mental foramen below P,_2 contact or below anterior root of P2; depth of ramus relatively constant from back to front, part below C and incisors deep and splayed outward; angular process of dentary prominent and strongly inflected; upper and lower incisors broad and spatulate; lowers splayed outward; I's separated from each other by small diastema; line of upper incisors semicircular (U-shaped); I1 larger than I2"5, I2"3 subequal in size and shape; I4 smaller and I5 smaller still; I,_4 subequal in size and shape; canines weakly to strongly developed; when well developed, canines set nearly perpendicular to jaw; when small, canines (especially lowers) incisiform-premolariform and tilted anteriorly; premolars increase sharply or gradually in length and breadth from P{ to P^; diastems may or may not separate lower premolars; P1 set at slight oblique angle in jaw; P3* has distinct posterobasal heel; P$ shorter and narrower than M\; in length M, < M2 > M3 > M4; in width M, < M2 > M3 > M4; on M,_3 trigonid cusps typically low, talonid broader than trigonid; talonid deeply basined; metaconid set directly linguad of protoconid; protoconid much larger than para- and metaconid, paraconid smaller than metaconid; hypoconulid ex- tremely median in position; M4 very reduced in size relative to Mj_3 and with 10 FIELDIANA: GEOLOGY talonid reduced to a nearly unbasined heel; in length M1 > M2 > M3 > M4; in width M1 < M2 2* M3 > M4; on M1"3 stylar shelf and cusps greatly reduced or absent, more so on M1 than M2"3; no distinct metastylar spur; parastyle small but distinct on M1"3; protocone large and talon distinctly basined; paracone and metacone large, low, subequal in size and height; former slightly smaller than latter on M1"2, subequal on M3; M4 extremely reduced in size; skull has large rounded braincase and short rostrum; interorbital region broad; broadest point of palate at anterior edge of M3; posterior nares broad and mesial septum con- tinues backward as narrow ridge along base of skull; palatal vacuities large and well-developed; premaxillaries greatly elongated resulting in notable develop- ment of precanine region of rostrum such that C is in relatively more posterior position in tooth row than occurs in other didelphoids; nasals expanded pos- teriorly; paroccipital process absent; postglenoid process of squamosal small, extremely short, and slightly concave anteriorly; ear region and ossicles highly specialized; auditory bullae large, ossified, much inflated, and occupy a relatively large area on base of skull; anterior one-third of bullar wall formed by large tympanic process of alisphenoid, and posterior two-thirds formed by inflated entotympanic (fide Segall, 1969b)4; ectotympanic does not participate in for- mation of bulla but is enclosed by bulla and forms a recess meatus only; ecto- tympanic thus hidden from lateral view; no open space on ventral side of bullae but a medial oval-shaped area is located in anteromedial part of entotympanic; tympanic cavity divided by two septa into three compartments, anterior, medial, and posterior; malleus differs greatly from other marsupials (Segall, 1969a, b); manubrial lever is elongated and lever ratio is greater than in other marsupials (Segall, 1969a, pp. 180-81); posterior carotid foramen appears lacking; pes di- dactylous; female with well-developed pouch or marsupium with four symmet- rically placed mammae; ossified epipubic bones present; diploid chromosome number 2n = 14. Known range. — Microbiotheres are, and according to the known fossil record always have been, endemic to South America. Fossils are known from beds of Colhuehuapian and Santacrucian age in Patagonia, southern Argentina. Living forms occur on Chiloe Island, south central Chile, and adjacent parts of Argen- tina. Microbiotherium Ameghino, 1887, p. 6. Microbiotherium Ameghino, 1887, p. 6. Stylognathus Ameghino, 1891, p. 309. 4 Reig and Simpson (1972, p. 525) believe that the identity of what Segall (1969 a,b, 1970) calls the entotympanic needs to be substantiated. Furthermore, Patterson (1965) believed that in addition to the alisphenoid and the inflated tympanic wing of the pars petrosa ( = entotympanic of Segall) contributing to formation of the bulla, that an inflated wing of the pars mastoidea was involved posteriorly. In support of his interpretation, Segall (1970, p. 197) noted that an entotympanic is found in several families of marsupials, as well as insectivores, and it increases in size with progressive specialization of the auditory region. Didelphis, he contends, has a small entotympanic and Dromiciops, a large one. The entotympanic in Dromiciops is, according to Segall (1969b, p. 500), separated on the outside from the tympanic process of the alisphenoid by a suture extending from the eustachian opening to the anterior part of the porus acusticus. This location, he believes, supports the assumption that the structure is an entotympanic and not a process of the periotic as Reig (1955a) and Patterson (1965) contend. MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 11 Eodidelphys Ameghino, 1891, p. 310. Prodidelphys Ameghino, 1891, p. 310. Hadrorhynchus Ameghino, 1891, p. 311; Ringuelet, 1953, p. 278 (as subgenus of Mi- crobiotherium). Phonocdromus Ameghino, 1894, p. 355. Proteodidelphys Ameghino, 1898, p. 187. Oligobiotherium Ameghino, 1902e, p. 124. Clenia5 Ameghino, 1904c, p. 260. Clenialites5 Ameghino, 1906, p. 422. Microbiotheridion Ringuelet, 1953, p. 280. Type of Microbiotherium. — M. patagonicum Ameghino, 1887, p. 6. Type of Stylognathus. — S. diprotodontoides Ameghino, 1891, p. 309. Type of Eodidelphys. — E. fortis Ameghino, 1891, p. 310. Type of Prodidelphys. — P. acicula Ameghino, 1891, p. 310. Type of Hadrorhynchus. — H. tortor Ameghino, 1891, p. 311. Type of Phonocdromus. — P. patagonicus Ameghino, 1894, p. 355. Type of Proteodidelphys. — P. praecursor Ameghino, 1898, p. 117, 187. Type of Oligobiotherium. — O. divisum Ameghino, 1902e, p. 124. Type of Clenia. — C. minuscula Ameghino, 1904c, p. 260. Type of Clenialites. — C. minusculus (Ameghino, 1904c, p. 260). Type of Microbiotheridion. — M. hernandezi (Simpson, 1932b, p. 2). Diagnosis. — Very small to large-sized microbiotheres; lower canine small to large, and set at or near perpendicular angle in jaw; P, set at slight to sharp oblique angle in jaw; posterobasal heel on P2 small but distinct; large mental foramina below P,_2 contact or below P2, and smaller foramen below M,; su- praorbital processes large and well-developed; para- and metaconules on M1-3 distinct but not large; middle cleft on M3 shallow or absent; bulla larger and more elongated anteriorly than in Dromiciops. Known Range. — Beds of Colhuehuapian and Santacrucian age, Chubut and Santa Cruz provinces, Patagonia, southern Argentina. Comments. — Included here within Ameghino's genus Microbiotherium are nine other genera proposed by him (i.e., Stylognathus, Eodidelphys, Prodidelphys, Ha- drorhynchus, Phonocdromus, Proteodidelphys, Oligobiotherium, Clenia, Clenialites) and a tenth (Microbiotheridion) proposed by Ringuelet. Ameghino's primary criteria for recognition of his genera included absolute overall size differences, relative size differences of the canine and paraconid, and structure of the mandibular symphysis. He also erroneously interpreted breakage patterns in the region of the lower canine and incisors as diagnostic features. The most extreme example of this was Ameghino's interpretation of the broken canine alveolus of the syntypes of Stylognathus as representing the alveolus of a large procumbent incisor. Ringuelet (1953) used the oblique orientation of P, and P2 in the jaw to diagnose Microbiotheridion. During the course of this study it became readily apparent that the characters used by Ameghino and Ringuelet to distinguish these taxa could, in many cases, be easily demonstrated to represent individual, age, or preservational differences among the various specimens which they studied. It became further apparent that, although many of these taxa could be readily relegated to synonymy, some 5 Ameghino (1906, p. 422; 1909, p. 206) substituted the name Clenialites for Clenia, which he said was preoccupied; however, Clenia is not preoccupied, and the name thus has priority by date of publication over Clenialites. 12 FIELDIANA: GEOLOGY could not. A number of discrete morphological types among the relatively large sample size could be defined, and these I here choose to recognize as warranting specific recognition. It is granted and demonstrated that size and morphological differences exist between these specimens and species, although taken together the spectrum of differences does not, in my judgment, warrant generic recog- nition. Consequently, I here recognize but a single fossil genus of microbiothere, Microbiotherium. Six species of Microbiotherium are here recognized: one (M. div- ision) is known from beds of Colhuehuapian age, four (M. acicula, M. patagonicum, M. tehuelchum, M. gallegosense) are known from the Santacrucian, and one (M. praecursor) is of uncertain age and provenance. Microbiotherium acicula (Ameghino, 1891). Figure 2; Table 2. Prodidelphys acicula Ameghino, 1891, p. 310; 1894, p. 361; 1898, p. 188. Eodidelphys acicula Ringuelet, 1953, p. 285, pi. 2, fig. 2. Prodidelphys pavita Ameghino, 1891, p. 310; 1894, p. 361; 1898, p. 188. Eodidelphys pavita Ringuelet, 1953, p. 286, pi. 2, fig. 3. Prodidelphys obtusa Ameghino, 1891, p. 311; 1894, p. 362; 1898, p. 188; 1906, p. 388, fig. 248. Eodidelphys obtusa Ringuelet, 1953, p. 287, pi. 2, fig. 4. Stylognathus diprotodontoides (partim, not lectotype) Ameghino, 1891, p. 309; 1894, p. 361. Microbiotherium patagonicum Sinclair, 1906 (partim), p. 415, pi. 62, figs. 5, 5a. Type of P. acicula. — MACN 5735, a fragment of a right mandibular ramus with alveoli of Ix_4 and M2, and C-Mj complete (listed as "tipo" in Ameghino's catalog and figured by Ringuelet, 1953, pi. 2, fig. 2). Type of P. pavita. — MACN 5736, a fragment of a left mandibular ramus with alveoli of Ij-C and Mj_2, and with Pt_3 complete (listed as "tipo" in Ameghino's catalog and figured by Ringuelet, 1953, pi. 2, fig. 3). Type of P. obtusa. — MACN 5737, a fragment of a left mandibular ramus with alveoli of Ii-P3, and C complete (listed as "tipo" in Ameghino's catalog and figured by Ameghino, 1906, fig. 248; and Ringuelet, 1953, pi. 2, fig. 4). Hypodigm. — The three types and MACN 5727, a left mandibular ramus with P3-M4 present and alveoli of rest of dentition (syntype of S. diprotodontoides and listed as "tipo" along with lectotype MACN 5728 of that species in Ameghino's catalog); and AMNH 9121, a fragment of a right mandibular ramus with alveoli of C-P2, and with P3-M2 present (figured by Sinclair, 1906, pi. 62, figs. 5, 5a). Horizon and locality. — All MACN specimens were collected in 1890-91 by Carlos Ameghino from the Santa Cruz Formation at Monte Observacion in Patagonia, southern Argentina; AMNH 9121 was collected by B. Brown in 1899 from near Felton's Estancia along the north bank of the Rio Gallegos. Age. — Santacrucian. Diagnosis. — Smallest known species of genus; averages 15% (n = 10, OR 0-30%) smaller in comparable linear tooth dimensions than M. patagonicum; lower C small, set nearly perpendicular in jaw, and about same height as M^ Pj set at slight oblique angle in jaw; small but distinct posterobasal heel on Pj and P2; P2 set straight in jaw; sometimes large diastem between Plx P2 and P3; posterobasal heel on P3 distinct but not large; slight size decrease from P3 to P^ P3 less than height of M,; M4 smaller (relative to M3) and talonid less well developed than in M. patagonicum and Dromiciops australis; P^ averages slightly longer than in M. patagonicum. MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 13 Fig. 2. Microbiotherium acicula (Ameghino, 1891, p. 310) (Santacrucian). MACN 5727, a left mandibular ramus with P3-M4: a, labial; b, occlusal; c, lingual views. Scale = 5 mm. Table 2. Measurements of lower cheek teeth of Microbiotherium acicula. Specimen AMNH 9121 MACN 5727 MACN 5735 P3 L W Ml L W M2 L W M3 L W M4 L W Ml-4 P3-M4 L L 1.0 0.8 0.5 0.5 0.5 1.3 1.3 1.2 0.9 1.0 1.0 1.4 1.4 1.0 1.0 1.2 1.0 1.0 0.6 4.9 5.7 Pl-3 L ca2.6 2.4 MACN 5736 1.0 0.6 Comments. — In 1891, Ameghino proposed the new microbiothere genus Pro- didelphys and assigned to it three species. The genotype, P. acicula, was char- acterized by its small size compared with other known microbiotheres. The type, MACN 5735, is a fragment of a right mandibular ramus with alveoli of lx„4 and M2, and with C-M, complete. The other two species, P. pavita and P. obtusa, are also of small size and were distinguished from P. acicula — and from each other — on the basis of relative size and proximity of the premolars and canine, and on the size and extent of the mandibular symphysis. The type of P. pavita, MACN 5736, is a fragment of a left mandibular ramus with alveoli of I,— C and M,_2/ and 14 FIELDIANA: GEOLOGY with Pj-3 complete. The type of P. obtusa, MACN 5737, is a fragment of a left mandibular ramus with alveoli of I,— P3r and with the C present. All three types are figured by Ringuelet (1953, pi. 2, figs. 2-4). The differences which exist between these types and the features which Ame- ghino used to differentiate them may all be attributable to individual and/or age differences of the specimens; they clearly are referrable to a single species for which the name acicula has priority by publication in a nominal list. A syntype (but not the lectotype) of Ameghino's (1891) taxon Stylognathus diprotodontoides (see p. 12) is also referrable to the species acicula. This specimen consists of a left ramus with P3-M4 present and with alveoli of rest of dentition. In addition, the specimen (AMNH 9121, a fragment of a right mandibular ramus with alveoli of C-P2 and with P3-M2 present) referred by Sinclair (1906, p. 415) to Microbiotherium patagonicum is also assignable to acicula. These two specimens provide important knowledge of structure of the molars of this species. Ringuelet (1953) recognized the species acicula, pavita, and obtusa as valid, and she placed them in the genus Eodidelphys (although she listed them in Prodidelphys in the caption to plate II). This assignment was made largely on the basis of the vertical orientation and small size of the canine, a feature shared with other taxa assigned to Eodidelphys. As discussed elsewhere (p. 11), I recognize Eodidelphys and Prodidelphys as junior synonyms of Microbiotherium and thus assign the species acicula to that taxon. Microbiotherium acicula is the name applied to this smallest known member of the genus. Microbiotherium patagonicum Ameghino, 1887. Figures 3, 4; Table 3. Microbiotherium patagonicum Ameghino, 1887, p. 6; 1889, p. 264; 1894, p. 361; Sinclair, 1906 (partim), p. 415; Ringuelet, 1953, p. 274, pi. 1, fig. 4. Stylognathus diprotodontoides (partim) Ameghino, 1891, p. 309; 1894, p. 361. Halmarhiphus didelpoides (sic) (partim, not lectotype) Ameghino, 1891, p. 308. Parhalmarhiphus didelphoides (partim) Ameghino, 1903, pp. 157, 163, fig. 86. Halmarhiphus didelphoides (partim) Ameghino, 1894, p. 357; 1899, fig. 4; 1903, p. 157, fig. 80; 1906, p. 357, fig. 195; Simpson, 1930, p. 56. Garzonia annectens (partim, not type) Ameghino, 1891, p. 307. Parhalmarhiphus annectens (partim) Ameghino, 1894, p. 357. Halmarhiphus nanus (partim, not type) Ameghino, 1891, p. 308; 1894, p. 357. Hadrorhynchus conspicuus Ameghino, 1891, p. 311; 1894, p. 362. Microbiotherium conspicuum Sinclair, 1906, p. 450. Microbiotherium conspicuus Ringuelet, 1953, p. 279, pi. 1, fig. 6. Type of M. patagonicum. — MLP 11-30, a partial right mandibular ramus with alveoli of Px-2, part of anterior root of P3, and with Mj_4 present (figured by Ringuelet, 1953, pi. 1, fig. 4). Lectotype of S. diprotodontoides. — MACN 5728, a left mandibular ramus with alveoli of Ii-Mj, and M2_4 present (listed as "tipo" in Ameghino's catalog along with MACN 5727, a syntype). Type of H. conspicuus. — MACN 5740, a fragment of a left mandibular ramus with alveoli of Ij— Plr and with P2-Mj complete (listed as "tipo" in Ameghino's catalog and figured by Ringuelet, 1953, pi. 1, fig. 6). Hypodigm. — The three types and MACN 1, a fragment of a right mandibular ramus with alveoli of I,-C, and Pj-Mj complete (listed as "tipo" of M. tehuelchum in Ameghino's catalog); MACN 5704, a fragment of a right mandibular ramus with M3_4 complete (possible syntype of G. annectens); MACN 5718, a fragment of a left mandibular ramus with M2_3 complete, and roots of M4 (syntype of H. 15 16 FIELDIANA: GEOLOGY Fig. 4. Microbiotherium patagonicum Ameghino, 1887, p. 6 (Santacrucian). MACN 8471, a right maxillary fragment with M1 4: a, labial; b, occlusal; c, lingual views. Scale = 3 Table 3. Measurements of cheek teeth of Microbiotherium patagonicum. Ml-3 P3 Ml M2 M3 M4 Ml-4 Specimen L W L W L W L W L W L Upper Cheek Teeth MACN 8471 1.7 1.3 1.5 1.7 1.5 1.7 0.9 1.1 4.3 Lower Cheek Teeth MACN 1 0.8 0.8 1.5 1.1 MACN 5718 1.7 1.1 1.5 1.0 MACN 5719 1.2 0.7 MACN 5721 1.6 1.1 1.5 1.0 MACN 5728 1.7 1.1 1.5 1.0 1.3 0.6 ca6.0 MACN 5740 0.9 0.7 1.5 1.2 didelphoides and figured under that name by Ameghino, 1903, figs. 80, 86; 1906, fig. 195, the M4 was lost since it was figured by Ameghino); MACN 5719, a fragment of a left mandibular ramus with M4 (syntype of H. didelphoides); MACN 5721, a fragment of a left mandibular ramus with M2-3 complete, and alveoli of Mj and M4 (syntype of H. nanus); and MACN 8471, a right maxillary fragment with M1 4 (figured by Reig, 1955a, fig. 1A). MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 17 Horizon and locality. — All specimens are from the Santa Cruz Formation, Santa Cruz Province, Patagonia, southern Argentina, and their localities and dates of collection are as follows: the type, MLP 11-30 was collected by Carlos Ameghino from "las barrancas del no Santa Cruz" (Ameghino, 1889, p. 264); La Cueva MACN 8471 (1892-93); Sehuen MACN 5718, 5719 (1890-91); Santa Cruz MACN 1 (probably 1887); Monte Observacidn MACN 5704, 5721, 5728, 5740 (1890-91); all MACN specimens were collected by Carlos Ameghino. Age. — Santacrucian. Diagnosis. — Small-sized microbiothere; averages 15% larger than M. acicula and averages 16% smaller than M. tehuelchum; lower canine small and set nearly perpendicular in jaw; P, set at slight oblique angle in jaw; P2 set straight in jaw; posterobasal heel on P3 distinct but not large; slight size decrease from P3 to P^ P3 less than height of M,; para- and metaconules on M1"3 distinct but not large; protocone on M1"3 narrower anteroposteriorly than in M. tehuelchum; labial ridge on stylar shelf less distinct than in M. tehuelchum; middle cleft on M2 very shallow; middle cleft on M3 shallow; M4 larger (relative to M3) than in D. australis and smaller than in M. tehuelchum. Comments. — Ameghino briefly diagnosed Microbiotherium patagonicum gen. y sp. n. in 1887 (p. 6), and in 1889 (p. 264) he presented an expanded diagnosis as follows: Es este, a buen seguro, uno de los mamiferos mas pequenos que hasta ahora se conozcan; desgraciadamente me es conocido tan solo por la mitad izquierda de la mandibula inferior, en la que estan implantados los ultimos tres molares verdaderos [M2_4], existiendo en la parte anterior los alveolos de los demas dientes. El alv£olo del incisivo, aunque bastante destrozado, prueba que este diente era comprimido transversalmente como en los generos Abderites, Acdestis y Epanorthus; la protuberanria que forma el alveolo sobre el lado interno, prueba que la base del diente llegaba hasta debajo del pmj [Mi]. Los dos primeros premolares [P] _2] (de los que s61o se conservan los alveolos), eran sumamente pequenos. El premolar cuatro [Mi], era de tamano relativamente mas considerable y con dos raices distintas, como parece ser tambi£n el caso para los anteriores; todos estos dientes, a lo menos los posteriores, tienen la raiz posterior de tamano considerablemente mayor que la anterior. Los cuatro pre- molares [P,-M,] ocupan un espacio longitudinal de 4 milimetros. El m- [M2] es el diente de tamano mas considerable de toda la mandibula, alcan- zando un diametro anteroposterior de 0m0015. La corona, bastante elevada, es mas angosta adelante, mas ancha atris y con dos pares de tuberculos transversales, altos y puntiagudos, de los cuales los que constituyen el par anterior son de tamano un poco menor que los del par posterior. El my [M3] es de tamano un poco menor, de ancho mds igual adelante y atras y tambien con los dos pares de tuberculos puntiagudos, uno adelante y otro atras, formando casi dos crestas transversales separadas por un valle profundo. El m-j [M4] es de tamano mucho mas pequeno que el precedente, pero tambien con dos pares de tuberculos, aunque mas bajos y de cuspides mas romas. Los tres ver- daderos molares [M2_4] ocupan un espacio longitudinal de cerca de 4 milimetros. La rama horizontal de la mandibula es casi derecha y de un alto casi uniforme de s61o dos milimetros. Los agujeros mentonianos son en niimero de dos colocados: el anterior debajo de la parte posterior del pmy [P3] y el posterior debajo de la parte anterior del my [M2]. La mandibula, a la que le falta el incisivo y la parte posterior de la rama horizontal, cuando entera apenas debia tener unos 12 milimetros de largo. In this publication, and elsewhere, Ameghino labeled the seven cheek teeth of marsupials as (from front to back) Pm„ Pm2, Pm3, Pm4, M„ M2, M3. These same teeth are referred to here as Pi, P2, P3, M„ M2, M3, M4. Based on her exhaustive study of fossil didelphoids in the MLP, Ringuelet 18 FIELDIANA: GEOLOGY (1953, p. 274) deduced that MLP 11-30 was probably the specimen referred to by Ameghino and was thus the type of M. patagonicum. The only problem with this decision is that MLP 11-30 is a right mandibular ramus; Ameghino listed his specimen as a left. In all other respects the description and measurements presented by Ameghino accord with those of MLP 11-30 as published by Rin- guelet. It is almost certain that Ameghino's description of the type of M. patagonicum in his 1889 publication was made from either notes or memory, or a combination of these sources. F. Ameghino resigned from the MLP on 17 January 1888, the culmination of a progressive deterioration of relations with the director Alcedes Moreno, who thereupon barred Ameghino from the collections of that institu- tion. Ameghino began work on his 1889 monograph in February 1888. Moreno did not allow him access to the specimens in the MLP, and Ameghino specifically stated (1889, p. xiii-xiv) that Moreno was indeed responsible for the fact that a number of specimens listed and described in the 1889 paper were not figured. Further details of these problems and of ensuing consequences with regard to the whereabouts of various specimens collected by Carlos Ameghino in 1887 and deposited in the MLP are discussed elsewhere (Marshall, 1980, p. 36n). The important point is that an explanation exists for the potential error in Ameghino's referral to the type as a left ramus when in fact it is a right. Realizing these developments and considering the agreement in structure and size of MLP 11-30 with Ameghino's description of the type, it appears permissible to follow Rin- guelet and to regard MLP 11-30 as the type of M. patagonicum, despite the fact that it is not specifically labeled as such. I have not studied this specimen firsthand, although it is figured by Ringuelet (1953, pi. 1, fig. 4) in labial view. Ringuelet's measurements of MLP 11-30 are as follows: Longitud del borde anterior del alveolo del Pm, [PJ al borde posterior del M4 9.5 mm Longitud ocupada por los alveolos de los tres premolares 3.4 mm Longitud ocupada por los M!-M4 6.4 mm Altura de la mandfbula debajo del alveolo de Pm3 2.5 mm Altura de la mandfbula debajo de M4 3.5 mm In Ameghino's description of this species he records the presence of a large alveolus at the anterior end of the jaw which he regards as for an incisor and which he envisions further to be as large and procumbent as in species of the caenolestid genera Abderites, Acdestis, and Epanorthus ( = Palaeothentes, see Mar- shall, 1980). As shown in Ringuelet's figure of this specimen (pi. 1, fig. 4), however, it is broken anterior to the P! and the partial alveolus retained at the anterior-most edge is that of the canine. The breakage pattern is such that it permits the interpretation reached by Ameghino, but knowledge of other spec- imens referable to this and related species clearly indicates the true identity of this alveolus. Several of Ameghino's other species are in whole or in part referable to M. patagonicum. Stylognathus diprotodontoides was erected by Ameghino (1891, p. 309) on two relatively complete left mandibular rami: MACN 5727, with P3-M4 present and alveoli of rest of dentition; and MACN 5728, with M2_4 present and alveoli of C-M,. These specimens are syntypes and both are listed as "tipo" in Ame- ghino's catalog. The first is here referred to Microbiotherium acicula (see p. 12), and the second is relegated to lectotype of S. diprotodontoides. MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 19 Ameghino (1891, p. 309) noted that S. diprotodontoides "[s]olo tiene dos inci- sivos inferiores muy pequenos, el canino es regularmente desarrollado; . . ." Both MACN 5727 and 5728 are missing the anterior-most edge of the mandibular rami such that, in the former, only two incisor alveoli are preserved and, in the latter, only one. Thus, the presence of only two incisors as noted by Ameghino is a feature attributable to this breakage. Judging from the canine alveolus in both specimens, the tooth was moderately developed and was either oriented perpendicular in the jaw or had only a slight anterior inclination. Stylognathus is here (see p. 11) recognized as a junior synonym of Microbiotherium and the species diprotodontoides is regarded as a junior synonym of M. patagonicum. Halmarhiphus didelphoides Ameghino (1891, p. 308) was based on four speci- mens which can be regarded as syntypes (Reig, 1955b, p. 63). MACN 5716, a fragment of a right mandibular ramus with M,_2 present and roots of M3_4 (the M2 is figured by Ameghino, 1899, fig. 4), is here selected lectotype and is referred to Microbiotherium tehuelchum (see p. 26). MACN 5717, a left mandibular ramus with P3 and M2_3 present and roots of M, and M4, is referable to Stilotheriutn dissimile (see p. 67). MACN 5718, a fragment of a left mandibular ramus with M2-3 complete and roots of M4 (figured by Ameghino, 1903, figs. 80, 86; 1906, fig. 195), and MACN 5719, a fragment of a left mandibular ramus with M4, are referable to M. patagonicum. Reig (1955b, p. 63) noted that according to Ameghino's catalog and the original species description, Garzonia annectens Ameghino (1891, p. 307) is the genotype of Parhalmarhiphus Ameghino (1894, p. 356). The type, MACN 5703, is a fragment of a left mandibular ramus with M, _4 and is referable to Stilotherium dissimile (see p. 66). The generic diagnosis of Parhalmarhiphus, however, appears to have been based on two other specimens; MACN 5704, a fragment of a right mandibular ramus with M3_4 complete, is referable to M. patagonicum, and MACN 5705, a fragment of a left mandibular ramus with M3_4 missing paraconid of M3, is an indeterminate caenolestid. Halmarhiphus nanus Ameghino (1891, p. 308) was based on three specimens (see Reig, 1955b, p. 63). The type, MACN 5720, a fragment of a right mandibular ramus with alveoli of M,, M4, and M23 present, is referable to Stilotherium dissimile (see p. 66). MACN 5721, a fragment of a left mandibular ramus with M2_3 complete and alveoli of M, and M4, is here referred to M. patagonicum. The third specimen, MACN 5722, a fragment of a left mandibular ramus with P3 present and alveoli of P,_2 and M,_4, is referable to Phonocdromus gracilis (see p. 69; Reig, 1955b, p. 63). Hadrorhynchus conspicuus Ameghino (1891, p. 311) was based on MACN 5740, a fragment of a left mandibular ramus with alveoli of Ii— Pi and with P^M, complete. This specimen, listed as "tipo" in Ameghino's catalog, is figured by Ringuelet (1953, pi. 1, fig. 6) and is here referred to M. patagonicum. Microbiotherium praecursor (Ameghino, 1898). Figure 5. Proteodidelphys praecursor Ameghino, 1898, p. 117, 187, fig. 52b; 1899, p. 557, figs. 1, 2, 15; 1900, p. 201, figs. 2-5; 1902a, pp. 6-7; 1902b, p. 21, figs. 1-3; 1902c, p. 421, figs. 1, 2, 15; 1903, p. 161, figs. 82, 84, 88; 1904a, p. 53, figs. 50-51; 1904b, p. 35, fig. 21; 1906, p. 288, fig. 69; Simpson, 1929, p. 130; 1932a, p. 4, fig. 1A; 1938, p. 3. Eodidelphys praecursor Ringuelet, 1953, p. 284. Type of P. praecursor. — MACN 10799, a right mandibular ramus with alveoli of L._2 and I3-M4 complete (posterior edge of ramus and I,_2 have been lost since -Q O © jy in p 20 MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 21 specimen was figured by Ameghino, 1898, fig. 52b; 1899, figs. 1, 2, 15; 1900, figs. 2-5; 1902b, figs. 1-3; 1902c, figs. 1, 2, 15; 1903, figs. 82, 84, 88; 1904a, figs. 50, 51; 1904b, fig. 21; 1906, fig. 69); I2 has been lost since specimen was figured by Simpson (1932a, p. 4, fig. 1A). Hypodigm. — Type only. Horizon and locality. — "Cretaceo inferior (areniscas abigarradas) de Patagonia" (fide Ameghino, 1898, p. 117). Collected by Nicolas Illin, an employee of the Museo de La Plata (Simpson, 1932a, p. 6). Age. — Unknown (see below). Measurements. — P3 L = l.l mm, W = 0.8 mm; M[ L = 1.7 mm, W = 1.2 mm; M2 L— 1.7 mm, W = 1.2 mm; M3 L = 1.7 mm, W— 1.1 mm; M» L = 1.3 mm, W = 0.8 mm; M,_„ = 6.8 mm; Pj-M4 = 8.0 mm. Diagnosis. — Indistinguishable in size and structure from smaller specimens of the Santacrucian species Microbiotherium tehuelchum (see below); C appears to be smaller than in those specimens. Description. — Dental formula Fr, Cy, Pj, M7. Incisors subequal, spaced, spat- ulate. Canine erect, root single but grooved, somewhat recurved crown, inner side excavated, very slight rudiment of internal heel. Premolars progressively larger, Pj to P3. Molars of generalized didelphid pattern, trigonid little elevated above talonid. Protoconid larger than metaconid, metaconid larger than para- conid. Talonid typically didelphid on M,_3. M,^ subequal, M3 somewhat smaller, and M4 very small and with greatly reduced talonid, shorter and narrower than trigonid and with only one really distinct cusp. Horizontal ramus of moderate proportions, symphysis extending to between canine and P„ fused, mental foramina under P,_2 and anterior end of M, (Simpson, 1932a, p. 4). Comments. — As demonstrated by Simpson (1932a), the location and age of the beds from which Proteodidelphys praecursor was collected are not known. This species was included by Ameghino (op. cit.), along with four isolated mammal teeth and "a small hard ball of doubtful nature" (Simpson, 1932a, p. 3), as part of his "Proteodidelphys Fauna." Ameghino (e.g., 1900 and elsewhere) noted that these specimens came from the "areniscas abigarradas" or from what is now called the Cretaceous dinosaur-bearing Chubutiano beds from the lower Chubut Valley in Patagonia. Simpson (1932a, p. 6) noted that Ameghino gave no data in support of this claim, despite the fact that he made frequent reference to these specimens. On the contrary, Simpson (ibid.) amassed considerable evi- dence, albeit some circumstantial, which suggests that these fossils are of Tertiary age and may have come from beds of Casamayoran, Mustersan, Deseadan, and/ or Colhuehuapian age, all of which occur in or near the lower Chubut Valley and all of which may or do contain mammals. Most importantly, the specimens were not found by Carlos Ameghino, but by a collector whose field data is not reliable. Scott (1937) compromised the argument by assigning P. praecursor to the Riochican, but this is contrary to all the evidence and solves none of the un- answered problems (Simpson, 1948, p. 35, n. 1). Simpson further noted (ibid., p. 5): There is no way in which Proteodidelphys can be distinguished generically from the common Eodidelphys. Certain specimens from the Santa Cruz formation referred by Ameghino to Eodidelphys famula cannot surely be distinguished specifically: size, struc- ture, and position of each tooth, depth and form of mandible, position and size of mental foramina, even the non-essential characters of color and mode of preservation, are all nearly identical. 22 FIELDIANA: GEOLOGY Technically it would perhaps be necessary to reduce the names of this genus and species to synonymy, as they cannot be defined in a diagnostic way, but in view of their history and the doubt that still clouds their true significance, it seems to be practical and permissible to retain them tentatively. Based on Simpson's comments, Ringuelet (1953, p. 284) referred the species praecursor to the genus Eodidelphys. I have compared praecursor with most microbiotheres discussed or listed in this study and find that it is morphologically indistinguishable from specimens of the Colhuehuapian species Microbiotherium divisum and the Santacrucian spe- cies M. tehuelchum (in which E. famula is here included as a junior synonym). Furthermore, praecursor is slightly smaller than all of the known specimens of M. divisum, but does overlap in size with the smaller specimens of the M. tehuelchum sample. In this study I follow Simpson's caution and tentatively retain praecursor as a distinct species, although I do assign it to Microbiotherium (to include Eodidelphys as a junior synonym), as it is indistinguishable from members of that taxon. Microbiotherium divisum (Ameghino, 1902e). Figures 6, 7; Table 4. Oligobiotherium divisus (sic) Ameghino, 1902e, p. 124; Ringuelet, 1953, p. 268. Oligobiotherium divisum Simpson, 1930, p. 36; Reig, 1955a, p. 121. ? Microbiotherium divisus (sic) Pascual and Odreman Rivas, 1971, p. 387. Clenia minuscula Ameghino, 1904c, p. 260; Ringuelet, 1953, p. 281; Reig, 1955a, p. 121 (as junior synonym of O. divisum). Clenialites minusculus Ameghino, 1906, p. 422n, fig. 315; 1907, fig. 1; Reig, 1955a, p. 122 (as junior synonym of O. divisum). Microbiotherium hernandezi Simpson, 1932b, p. 2, fig. 1; Pascual and Odreman Rivas, 1971, p. 387. Microbiotheridion hernandezi Ringuelet, 1953, p. 280; Reig, 1955a, p. 122 (as junior synonym of O. divisum). Type of Oligobiotherium divisus. — MACN 52-374, a fragment of a right mandib- ular ramus with posterior root of M2, M3 missing metaconid, and roots of M4. Type of Clenia minuscula. — MACN 52-381, a fragment of a left mandibular ramus with M,_2 (figured by Ameghino, 1906, fig. 315; 1907, fig. 1). Type of Microbiotherium hernandezi. — AMNH 29664, a right mandibular ramus with crowns of P2-M4 (protoconid, paraconid, and hypoconid of M, are partially broken), and alveoli of other teeth (figured by Simpson, 1932b, fig. 1). Hypodigm. — The three types and AMNH 29684, a fragment of a right man- dibular ramus with P3-M2 complete; AMNH 29708, a fragment of a right man- dibular ramus with alveoli of P2 and M3, and with P3-M2 complete; MMP M-217, a fragment of a right mandibular ramus with alveoli of P,_3, Mj complete, and roots of M2; MMP M-942, a fragment of a left mandibular ramus with alveoli of I,-C, P,_3 complete, and alveoli of M,; MMP M-943, a fragment of a left man- dibular ramus with roots of C-P2/ and P3-M2 complete; MMP M-951a, a nearly complete edentulous left mandibular ramus; MMP M-951b, a fragment of a right mandibular ramus with alveoli of M]_4 (probably same individual as MMP M-951a). Horizon and locality. — Colhue-Huapi beds at the Barranca south of Lago Colhue- Huapi, southern Argentina. The MACN specimens were collected by Carlos Ameghino; the AMNH specimens were collected by members of the Scarritt Expedition to Patagonia in 1930; MMP M-217 was collected by G. Scaglia in 1950; Fig. 6. Microbiotherium division (Ameghino, 1902e, p. 124) (Colhuehuapian). MACN 52-381 (type of Clenia minuscula), a fragment of a left mandibular ramus with M,-2: a, labial; b, occlusal; c, lingual views. Scale = 3 mm. Table 4. Measurements of lower cheek teeth of Microbiotherium divisum. Ml M2 M3 M4 Ml-4 P1-M4 Pl-3 Specimen L W L W L W L V\ ' L L L MACN 52-374 2.0 ca 1.2 MACN 52-381 1.8 1.3 1.8 1.3 AMNH 29664» 2.0 1.3 2.0 1.4 2 1 1.3 1.5 0.9 ca 11.7 AMNH 29684 1.8 1.2 1.9 1.3 AMNH 29708 1.9 1.3 2.0 1.4 MMP M-217 1.8 1.3 MMP M-942 4.4 MMP M-943 1.8 1.3 1.9 1.4 MMP M-951a ca7.6 ca 11.7 MMP M-951b 7.8 * P3L = 1.4 mm, W = 0.9 mm. 23 24 MARSHALL: SYSTEMAT1CS OF MICROBIOTHERIIDAE 25 the other MMP specimens were collected by G. Scaglia, Contreras and J. Her- nandez in 1964 (MMP specimens are from "frente a Estacion La Parada"). Age. — Colhuehuapian. Diagnosis. — Medium-sized microbiothere and only known microbiothere of Colhuehuapian age; similar in size and structure to larger specimens of Santa- crucian species M. tehuelchum; lower canine medium to large size and set nearly perpendicular in jaw; P, set at slight oblique angle in jaw; P2 set straight in jaw; posterobasal heel on P2 small but distinct; posterobasal heel on P3 distinct but not large; slight size decrease from P3 to P^ P3 about same height as M,. Comments. — Ameghino (1902e, p. 124) originally placed Oligobiotherium divisum in the family Microbiotheriidae, group Pedimana. He later (1904c, p. 260) named Clenia minuscula on the basis of a jaw fragment with two teeth from the same fauna and locality as Oligobiotherium divisum, and likewise placed it in the family Microbiotheriidae. In the original description of C. minuscula, Ameghino com- pared it only with Pachybiotherium and made no mention of O. divisum. In 1906, Ameghino (p. 422n) substituted the name Clenialites for Clenia, which he said was preoccupied, although he did not indicate where that name had earlier been used. I was unable to find an earlier usage of the name Clenia; in a nominal list, the name Clenia has priority over Clenialites. Simpson (1932b, p. 2) erected another Colhuehuapian species of microbio- there, Microbiotherium hernandezi, on a nearly complete right mandibular ramus with most of the dentition, collected from the Barranca south of Lago Colhue- Huapi. He noted that M. hernandezi was much smaller than Pachybiotherium acclinum. The only other possible didelphid yet described from the Colpodon Beds is Oligo- biotherium divisum. This [species] nearly agrees with [M. hernandezi] ... in size, but Ameghino describes . . . [O. divisum] as having only two trigonid cusps on M3 and only one on M4, M3 with the talonid cusps in a straight transverse series, and M4 with a large basined talonid — characters so distinctive that it is doubtful whether Oligobio- therium belongs in this family at all. Ringuelet (1953, p. 280) placed Simpson's species hernandezi in a new genus, Microbiotheridion. Her primary basis for doing so was ... la posici6n de los premolares: mientras en los microbiotherinos el Pm, [PJ y el Pm2 [P2] estan colocados transversalmente con relaci6n al resto de la serie dentaria, en esta especie, dichos dientes estan colocados en linea con relaci6n a los restantes; siendo el unico resto atribuido al g£nero Microbiotherium que posee este caracter (Rin- guelet, 1953, p. 281). Based on an unpublished collaborative study with B. Patterson, Reig (1955a, p. 121) recognized these three species (divisum, minuscula, hernandezi) as synony- mous. I have examined all three types and concur with Reig's conclusion. The partial M4 was lost from the type of O. divisum since that specimen was described by Ameghino. His reference to the presence of only two trigonid cusps on the M3 (a feature commented upon by Simpson, 1932b) was based on the fact that the metaconid of that tooth was broken and missing. In all comparable size and structural features, the types and the referred specimens clearly are representative of a single species for which the name divisum has priority. The species divisum is the only known microbiothere of Colhuehuapian age. In size and structure divisum is very similar to the Santacrucian species Micro- biotherium tehuelchum, and divisum is here regarded as the direct Colhuehuapian 26 FIELDIANA: GEOLOGY ancestor of that species (see p. 52). Accordingly, divisum is placed in the genus Microbiotherium. Microbiotherium tehuelchum Ameghino, 1887. Figures 8-10; Tables 5-7. Microbiotherium tehuelchum Ameghino, 1887, p. 7; 1889, p. 265 (the specimen figured in pi. 1, fig. 17 as M. tehuelchum is a palaeothentine); 1891, p. 309; 1893, p. 78, fig. 2; 1894, p. 361, fig. 42; 1898, p. 187, fig. 52a; 1906, p. 386, fig. 245; Sinclair, 1906, p. 414, pi. 62, figs. 4, 4a, 6-12; Ringuelet, 1953, p. 275, pi. 1, fig. 1; Segall, 1969b, p. 495, fig. 5. Halmarhiphus didelpoides (sic) (partim) Ameghino, 1891, p. 308. Parhalmarhiphus didelphoides (partim) Ameghino, 1903, pp. 157, 163, fig. 86. Halmarhiphus didelphoides (partim) Ameghino, 1894, p. 357; 1899, fig. 4; 1903, p. 157, fig. 80; 1906, p. 357, fig. 195; Simpson, 1930, p. 56. Microbiotherium forticulum Ameghino, 1891, p. 309; 1894, p. 361; Ringuelet, 1953, p. 275, pi. 1, fig. 2 (as junior synonym of M. tehuelchum). Eodidelphys fortis Ameghino, 1891, p. 310; 1894, p. 361; 1905, p. 11, fig. 10; Ringuelet, 1953, p. 282, pi. 2, figs. 1, 5, 6; Reig, 1955a, p. 124, fig. 2A. Eodidelphys famula Ameghino, 1891, p. 310; 1894, p. 361; 1898, p. 187; 1900, p. 201; 1905, p. 11, fig. 11; 1906, p. 387, fig. 246; Ringuelet, 1953, p. 283. Phonocdromus patagonicus Ameghino, 1894, p. 355 (partim); Reig, 1955b, p. 64 (invalid). Hadrorhynchus tortor Ameghino, 1891, p. 311; 1894, p. 362. Microbiotherium tortor Sinclair, 1906, p. 412, pi. 62, figs. 1, 2, 2a; Ringuelet, 1953, p. 278, pi. 1, fig. 3. Hadrorhynchus torvus Ameghino, 1891, p. 311; 1894, p. 361. Microbiotherium torvus Sinclair, 1906, p. 412 (as junior synonym of Microbiotherium tortor); Ringuelet, 1953, p. 280, pi. 1, fig. 5. Neotype of M. tehuelchum. — MLP 11-36, a nearly complete right mandibular ramus with P2-M4 complete (figured by Ringuelet, 1953, pi. 1, fig. 1; Pascual & Herrera, 1975, pi. 1, fig. 1). Lectotype of H. didelphoides. — MACN 5716, a fragment of a right mandibular ramus with NV2 complete, and roots of M3_4 (listed as "tipo" in Ameghino's catalog; figured by Ameghino, 1899, fig. 4). Lectotype ofM. forticulum. — MACN 5725, a fragment of a left mandibular ramus with alveoli of C and M2, and Pj-Mj complete (listed as "tipo" in Ameghino's catalog; figured by Ringuelet, 1953, pi. 1, fig. 2). Lectotype of E. fortis. — MACN 5729, a left mandibular ramus with alveoli of It-C, and Pj-M, complete (listed as "tipo" in Ameghino's catalog; figured by Reig, 1955a, fig. 2A). Lectotype ofE. famula. — MACN 5732, a right mandibular ramus with Pi-M4 and alveoli of all other teeth (listed as "tipo" in Ameghino's catalog). Lectotype of P. patagonicus. — MACN 5741, a left mandibular ramus with partial dentition. Type of H. tortor. — MACN 5738, a fragment of a right mandibular ramus with alveoli of Ij-C, roots of P,_2, P3 complete, and alveoli of M,_2 (listed as "tipo" in Ameghino's catalog). Type ofH. torvus. — MACN 5739, a fragment of a right mandibular ramus with alveoli of C, roots of Pu P2_3 complete, and alveoli of M, (listed as "tipo" in Ameghino's catalog; figured by Ringuelet, 1953, pi. 1, fig. 51). Hypodigm. — The eight types and MACN 5724, a fragment of a left mandibular ramus with I3-P, and P3 present, and alveoli of I,_2, P2 and M^ (labeled M. tehuelchum and figured under that name by Ameghino, 1893, fig. 2; 1894, fig. 42; 1898, fig. 52a; 1906, fig. 245; P2 and Mj have been lost since the specimen MARSHALL: SYSTEMATICS OF MICROBIOTHER1IDAE 27 Fig. 8. Microbiotherium tehuelchum Ameghino, 1887, p. 7 (Santacrucian). PU 15038, a nearly complete left mandibular ramus with alveoli of Ii_4, C-M3 complete, and roots of M4: a, labial; b, occlusal; c, lingual views. Scale = 10 mm. was figured); MACN 5726, a fragment of a left mandibular ramus with alveoli of C and Mu roots of P,-2/ and P3 complete (labeled M. forticulum); MACN 5730, a left mandibular ramus with alveoli of C-P2 and M2, and P-j-M, and M3 4 com- plete (probable syntype of E. fortis); MACN 5731, a left mandibular ramus with M! and M3 complete, and alveoli of rest of dentition except M4 (probable syntype of E. fortis); MACN 5733, a left mandibular ramus with P2-M3 complete (probable syntype of E. famula and the specimen measured in the original description of that species); MACN 5734, a fragment of a left mandibular ramus with alveoli of I,_4 and M„ and C-P3 present (probable syntype of £. famula); MACN 8472, a fragment of a right mandibular ramus with alveoli of I,-C, and P, . 2 complete (labeled M. tehuelchum); MACN 8474, a right mandibular ramus with alveoli of 1,-P,, and P^M-, complete; MACN 8475, a left mandibular ramus with alveoli of Pi_2/ and P3-M4 complete; MACN 8476, a fragment of a left mandibular ramus with alveoli of C, Pj_3 complete, anterior root of M„ talonid of M, present, and roots of M2 (labeled Stylognathus crassus); MACN 8477, a fragment of a right 28 MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 29 Table 5. Measurements of upper cheek teeth of Microbiotherium tehuelchum. Ml M2 M3 M4 Ml-3 Specimen L W L W L W L W L MACN8505 2.0 1.6 2.0 1.9 1.7 1.9 0.9 1.2 5.4 MLP 58-IX-3-4(l) 1.9 1.7 2.0 2.0 1.8 2.0 5.4 MLP 58-IX-3-4(r) 1.9 1.7 1.9 2.0 1.8 2.0 0.9 1.5 5.4 PU 15038 1.9 2.2 1.8 2.1 PU 15698(1) 2.2 2.2 2.2 2.3 2.0 6.1 Table 6. Measurements of lower cheek teeth of Microbiotherium tehuelchum. F M, M2 M3 M4 M,.4 P*-M4 Specimen L W L W L W L W L W L L MACN 5716 2.0 1.3 2.0 1.3 ca7.2 MACN 5724 1.1 0.7 MACN 5725 1.3 0.8 1.9 1.3 MACN 5726 1.2 0.8 MACN 5729 1.3 0.7 1.9 1.3 1.9 1.3 1.8 1.2 1.5 0.8 7.4 8.8 MACN 5730 1.3 0.8 1.9 1.2 1.9 1.3 1.4 0.8 7.6 9.0 MACN 5731 2.0 1.3 1.8 1.2 MACN 5732 1.3 0.7 1.8 1.2 1.8 1.3 1.7 1.1 1.3 0.8 7.5 8.8 MACN 5733 1.1 0.8 1.9 1.3 1.9 1.3 1.8 1.1 MACN 5734 1.3 0.8 MACN 5738 1.3 0.9 MACN 5739 1.3 0.9 MACN 8474 1.1 0.8 1.7 1.3 1.7 1.3 MACN 8475 1.1 0.7 1.7 1.2 1.7 1.3 1.7 1.1 1.3 0.8 6.8 7.8 MACN 8476 1.1 0.7 1.3 MACN 8477 1.1 0.8 1.7 1.2 1.7 1.2 MACN 8478 1.7 1.1 MACN 8479 1.3 0.9 1.8 1.2 1.8 1.3 MACN 8480 1.3 0.7 1.9 1.3 1.9 1.3 1.8 1.1 1.4 0.9 7.3 8.7 MACN 8481 1.3 0.8 1.8 1.4 1.8 1.3 1.4 1.0 MACN 8483 1.2 0.7 1.9 1.3 1.8 1.3 1.8 1.2 1.4 0.8 7.1 8.3 MACN 8503 1.9 1.1 MACN 8504 1.9 1.2 1.7 1.1 1.4 0.8 MACN 8506 1.1 0.7 1.7 1.1 1.7 1.2 1.6 1.1 1.1 0.8 6.4 7.6 MACN 8507 1.7 1.2 1.6 1.1 1.2 0.8 PU 15038 1.2 0.8 1.9 1.2 1.8 1.4 1.8 1.2 ca7.0 ca8.5 PU 15698(r) 1.3 0.9 2.0 1.4 2.0 1.5 1.9 1.3 1.5 1.1 7.7 9.2 PU 15698(1) 1.3 0.8 2.0 1.3 1.9 1.4 1.8 1.3 ca7.6 9.2 mandibular ramus with P2-M2 complete; MACN 8478, a fragment of a left man- dibular ramus with alveoli of Pi-M„ and M2 complete; MACN 8479, a right mandibular ramus with alveoli of Ii-C, and Pi-M2 complete (labeled £. fortis and figured by Ringuelet, 1953, pi. 1, fig. 5); MACN 8480, a right mandibular ramus with P3-M4 complete (labeled E. fortis and figured by Ringuelet, 1953, pi. 1, fig. 1 erroneously as MACN 8479); MACN 8481, a left mandibular ramus with alveoli of I,-C, P13 and M2 4 complete, and trigonid of M, missing (labeled £. fortis and figured by Ringuelet, 1953, pi. 1, fig. 6); MACN 8483, a left mandibular ramus with P2-M4 complete, and alveoli of rest of dentition (labeled E. famula); MACN 8493, a calcaneum (associated with MACN 8483 and figured as E. famula by Ameghino, 1905, fig. 11); MACN 8503, a left mandibular ramus with P, and M3 complete, and alveoli of rest of dentition (labeled E. famula); MACN 8504, a 30 FIELDIANA: GEOLOGY ANTERIOR COMPARTMEN SUTURE _OVAL SHAPED AREAfMIDDLE COMPARTMENT) OSTERIOR COMPARTMENT '-.BREAKAGE SEPTUM 6 mm Fig. 10. Microbiotherium tehuelchum Ameghino, 1887, p. 7 (Santacrucian). PU 15038, ventral view of left auditory region (after Segall, 1969b, fig. 5). fragment of a left mandibular ramus with M2_4 complete (labeled £. famula and figured by Ameghino, 1906, fig. 246e, o); MACN 8505, a very crushed skull with left M1"4 complete (figured as E. famula by Ameghino, 1906, fig. 246a); MACN 8506, a right mandibular ramus with P,-M4 complete (labeled £. famula); MACN 8507, a right mandibular ramus with M2_4 complete (labeled £. famula); MACN 10248, an edentulous right mandibular ramus with alveoli of all teeth (labeled £. famula); MACN ?, an astragalus (figured as £. fortis by Ameghino, 1905, fig. 10); PU 15038, a left maxillary fragment with M2"3 complete; a nearly complete left mandibular ramus with alveoli of 1^4, C-M3 complete, and roots of M4; part of a scapula, humerus, radius, ulna, cervical vertebrae, and the greater part of the left bullar region of skull (figured by Sinclair, 1906, pi. 62, figs. 4, 4a, 6-12; Segall, 1969b, fig. 5); PU 15698, a partial skull with alveoli or roots of right side; alveoli of p-2-4"5; complete P, C, and M1"2; and roots of P1 and parts of crowns of M3"4 on right side; part of right mandibular ramus with P3-M4 present; and greater part of left mandibular ramus with Ij_2 present, roots of I3_4, C-M3 present, roots of M4 (figured by Sinclair, 1906, pi. 62, figs. 1, 2, 2a); MLP 11-35, a fragment of a left mandibular ramus with P2-M3 present (figured by Ringuelet, 1953, pi. 1, fig. 3); MLP 58-LX-3-4, rostral portion of skull with palate, roots of left C-P3 and M4, and with M1"3 complete, and with base of right C, roots of P2, MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 31 Table 7. Statistics for some cheek tooth dimensions of Microbiotherium tehuelchum. Dimension N OR X s cv Upper Cheek Teeth M1 L 4 1.9-2.2 2.0 0.14 7.00 W 4 1.6-2.2 1.8 0.27 15.00 M2 L 5 1.9-2.2 2.0 0.12 6.00 W 5 1.9-2.3 2.08 0.16 7.70 M3 L 5 1.7-2.0 1.82 0.11 6.04 W 4 1.9-2.1 2.0 0.08 4.00 M4 L 2 0.9 W 2 1.2-1.5 1.35 0.21 15.56 M1-3 4 5.4-6.1 5.58 0.35 6.27 Lower Cheek Teeth P3 L 22 1.1-1.3 1.22 0.09 7.38 W 22 0.7-0.9 0.78 0.07 8.97 M, L 18 1.7-2.0 1.86 0.11 5.91 W 19 1.1-1.4 1.25 0.08 6.40 M2 L 17 1.7-2.0 1.82 0.10 5.49 W 17 1.2-1.5 1.31 0.08 6.11 M3 L 17 1.6-1.9 1.77 0.09 5.08 W 17 1.1-1.3 1.17 0.08 6.84 M< L 11 1.1-1.5 1.35 0.12 8.89 W 11 0.8-1.1 0.85 0.10 11.76 M,_4 L 11 6.4-7.7 7.24 0.39 5.39 P3-M4 L 10 7.6-9.2 8.59 0.55 6.40 and with P1 and P3-M4 complete (figured as Microbiotherium sp. by Sinclair, 1906, fig. 6, p. 409). Horizon and locality. — All specimens are from the Santa Cruz Formation, Pa- tagonia, southern Argentina. The specific localities and dates of collection are as follows: Corriquen-Kaik MACN 8504, 8505 (collected in 1892-93); La Cueva MACN 8474, 8481, 8483, 10248 (collected in 1892-93); Quequa-Quemada MACN 8472, 8479, 8480, 8503 (collected in 1892-93); Monte Observation MACN 5716, 5724, 5725, 5726, 5729, 5730, 5731, 5732, 5733, 5734, 5738, 5739, 5741 (collected in 1890-91); Monte Observation MACN 8477, 8478 (collected in 1891-92); Monte Observation MACN 8475, 8476, 8506, 8507 (collected in 1892-93); Killik Aike Norte ( = Felton's Estancia) PU 15698 (collected by J. B. Hatcher and O. A. Peterson, 1896); PU 15038, MLP 11-35, MLP 11-36 and MLP 58-IX-3-4 are without specific locality data. The MACN specimens were collected by Carlos Ameghino. Age. — Santacrucian. Diagnosis. — Medium-sized microbiothere; averages 16% larger than M. vata- gonicum, and averages 29% smaller than M. gallegosense, and similar in size to larger specimens of M. divisum; lower canine medium to large and set nearly perpendicular in jaw; P, set at sharp oblique angle in jaw; lack of posterobasal heel on P,; P2 straight or set at slight oblique angle in jaw; posterobasal heel on P2 small but distinct; posterobasal heel on P3 large; very sharp size decrease from P3 to P„ P3 usually proodont and much larger than P2; bulla larger and more elongated anteriorly than in D. australis; supraorbital processes large and well developed; P1 and P2 set at slight oblique angle in jaw; P1 single-rooted (or nearly so) with no trace of posterobasal heel; P3 with large, well-developed posterobasal heel; P3 about same height as M„ para- and metaconules on M1"3 distinct but 32 FIELDIANA: GEOLOGY not large; M1"3 protocone broader anteroposteriorly and more robust than in M. patagonicutn or D. australis; stylar shelf with low but distinct labial ridge which encloses shelf lingually; middle cleft on M2 absent; middle cleft on M3 very shallow or absent; size of M4 (relative to M3) larger than in M. patagonicutn. Comments. — Microbiotherium tehuelchum was diagnosed briefly by Ameghino in 1887 (p. 7), and an expanded description and characterization was presented by him in 1889 (p. 265) as follows: Esta especie esta fundada en algunos pequenos fragmentos de mandfbulas infer- iores, el mas completo de los cuales es una rama izquierda conteniendo implantados parte del pmy, el pmy y los tres molares siguientes. Este animal se distingue de la especie anterior [M. patagonicum] por su tamano bastante mas considerable, por sus molares uno y dos de tamano mas igual entre si, y cada uno de ellos bastante mas angosto adelante que atras. Las muelas desde el pmy al my se presentan con una corona baja gastada por la masticacion, pero todavia aparentemente dividida en dos lobulos casi iguales y con las cuspides de los tuberculos casi completamente gastadas, menos las colocadas sobre el borde interno, que se conservan puntiagudas. Los dientes pmy y pmy, my y y parecen tener dos raices distintas, una anterior y otra posterior, pero el my, que es sumamente pequeno y de corona casi circular, no tiene probablemente mas que una sola raiz. Los dos ultimos premolares y los tres verdaderos molares, todos muy apretados unos a otros, ocupan un espacio longitudinal de 8 milimetros. La forma de la rama horizontal de la mandibula es, como en la especie precedente, con un alto uniforme de cerca de 3 milimetros, mostrando tambien en el lado externo dos pequenos agujeritos mentonianos, colocados: el primero, debajo de la parte media del pmy el segundo, debajo de la parte media del my; este ultimo agujerito se distingue del correspondiente en la especie precedente por su tamano excesivamente pequeno, apenas visible a simple vista. Ameghino further noted (ibid.) that the specimens were "[d]escubierta y col- lecionada ... en las barrancas del rio Santa Cruz, en Patagonia austral." He clearly stated that this taxon was erected on the basis of several small fragments of mandibular rami, of which the most complete was a left mandibular ramus with part of P3 [his pmy] and M:_4 [his "pmy y los tres molares"] present. His description and measurements of M. tehuelchum are clearly based in large part, if not totally, on this specimen. Unfortunately, no known specimens in the MLP collection agree perfectly with "the" specimen described by Ameghino. It would be appropriate to make this specimen the lectotype and this should be done when and if it is relocated. Several possibilities exist with regard to its whereabouts. Firstly, it may be mis- placed in the MLP collections or it could be lost. Secondly, Ameghino may have taken it with him to the MACN when he resigned from his post at the MLP. This had been shown to be the case for other type and referred specimens described by Ameghino in 1887 (see Marshall, 1980, p. 36n), and it could well be the case with M. tehuelchum. There is one specimen in the Ameghino collection in the MACN (MACN 1) which is labeled "tipo" of M. tehuelchum and is listed as being collected from Santa Cruz. MACN 1 is, however, not "the" specimen described by Ameghino (it is a fragment of a right mandibular ramus with alveoli of I]-C, and with Pi-Mj complete), but it may well be, and it probably is, one of the "algunos pequenos fragmentos de mandibular inferiores" to which he called attention in his 1889 diagnosis. To further complicate matters, MACN 1 is here referred to M. patagonicum on the basis that length of P3 (0.8 mm) and Mj (1.5 mm) (see table 3) is smaller and outside the range of the large sample MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 33 of specimens referred toM. tehuelchum (see tables 6, 7; L P3= 1.1-1.3 mm, N = 22; L M, = 1.7-2.0 mm, N = 18). Accordingly, MACN 1 is referred to the smaller- sized species M. patagonicum. The probability thus exists that among the original syntypes of M. tehuelchum some, as MACN 1, are referable to M. patagonicum. Despite these complications the measurements of M. tehuelchum presented by Ameghino do permit comparison with other specimens. Furthermore, as it is not possible to identify syntypes of M. tehuelchum in the MLP collection, it is necessary to select a neotype. I have thus chosen a specimen in the MLP, since that is the institutional collection where the original type was deposited. For neotype I select MLP 11-36, a nearly complete right mandibular ramus with P2-M4 complete (figured by Ringuelet, 1953, pi. 1, fig. 1; Pascual & Herrera, 1975, pi. 1, fig. 1). Ringuelet (1953, p. 275) believed MLP 11-36 to be a probable cotype of M. tehuelchum, and Pascual & Herrera (1975, pi. 1, fig. 1) figured it as type. This specimen, however, is more complete than the material described by Ame- ghino and it was therefore almost certainly not among the specimens referred to in his 1889 diagnosis of that species. Nevertheless, in size and structure MLP 11-36 agrees perfectly with Ameghino's diagnosis [I have not studied MLP 11-36 firsthand and its selection as neotype is based totally on my interpretation of the diagnosis given by Ameghino and Ringuelet]: e.g., L P3-M4 = 8.0 mm in both, while the depth of the ramus was listed as 3.8 mm by Ameghino and in MLP 11-36 is listed by Ringuelet as 3.5 mm below the P, and 4.0 mm below the M, and M4. Other measurements of MLP 11-36 given by Ringuelet (1953, p. 276) include: Longitud de la parte posterior del M4 a la parte anterior de la sinfisis 14 mm Longitud del espacio ocupado por Pm,-M4 10 mm Longitud del espacio ocupado por Pm,-Pm3 3.5 mm Longitud del espacio ocupado por M,-M4 7 mm Longitud de sinfisis 4 mm The specimen figured by Ameghino (1889, pi. 1, fig. 17) as M. tehuelchum is an unidentified member of the caenolestid subfamily Palaeothentinae. MACN 5724, a fragment of a left mandibular ramus with I3— P, and P3 present, and alveoli of I,_2/ P2 and M„ is referable to M. tehuelchum and was figured under that name by Ameghino (1893, fig. 2; 1894, fig. 42; 1898, fig. 52a; 1906, fig. 245). All spec- imens figured by Sinclair (1906) as M. tehuelchum (and M. tortor) are indeed referable to this species. As noted earlier (see p. 19), Halmarhiphus didelphoides Ameghino (1891, p. 308) was based on four specimens, of which MACN 5716 is here selected lec- totype and is referred to M. tehuelchum. The M2 of this specimen is figured by Ameghino (1899, fig. 4). The identity of the specimen from which the left M2 figured in occlusal view by Ameghino (1903, fig. 86) as Parhalmarhiphus didel- phoides is not surely known. The identities of the other three syntypes are dis- cussed on p. 19. Microbiotherium forticulum Ameghino (1891, p. 309) was erected on two spec- imens labeled "tipo" in Ameghino's catalog and which can be regarded as syntypes. The first, MACN 5725, is a fragment of a left mandibular ramus with alveoli of C and M2, and with P^M, complete. This is the more complete of the two specimens and clearly is the one measured by Ameghino. It is here selected lectotype and is figured by Ringuelet (1953, pi. 1, fig. 2). The second, MACN 34 FIELDIANA: GEOLOGY 5726, is a fragment of a left mandibular ramus with alveoli of C and M„ roots of Pj -2/ and P3 complete. Both specimens are discussed by Ringuelet (1953, p. 276) and referred by her to M. tehuelchum. I agree with this assignment. Eodidelphys fortis Ameghino (1891, p. 310) was, according to Ameghino's cat- alog, erected on the basis of three specimens. MACN 5729, a left mandibular ramus with alveoli of Ij-C and Pt-M4 complete, is listed as "tipo" in Ameghino's catalog; it is here selected lectotype and is figured by Reig (1955a, fig. 2A). Two additional specimens, MACN 5730, a left mandibular ramus with alveoli of C-P2 and M2, and P3-M, and M3_4 complete; and MACN 5731, a left mandibular ramus with M, and M3 complete, and alveoli of rest of dentition except M4, are both labeled £. fortis. All of these specimens are referred confidently here to M. tehuelchum. I could not locate the astragalus figured by Ameghino (1905, fig. 10) as E. fortis. Ringuelet (1953, p. 282) described and figured three specimens as £. fortis: MACN 8479 (pi. 1, fig. 5), 8480 (pi. 1, fig. 1— listed erroneously as 8479), and 8481 (pi. 1, fig. 6); all are referable to M. tehuelchum. Eodidelphys famula Ameghino (1891, p. 310) was apparently erected on the basis of three specimens. MACN 5732, a right mandibular ramus with Pi-M4 and alveoli of all other teeth is listed as "tipo" in Ameghino's catalog and is here selected lectotype. MACN 5733, a left mandibular ramus with P2-M3 complete, is most probably the specimen measured by Ameghino in the original description and is labeled E. famula in Ameghino's catalog. The third specimen, MACN 5734, a fragment of a left mandibular ramus with alveoli of Ii_4 and Mlr and with C-P3 present, is also labeled E. famula. The calcaneum figured by Ameghino (1905, fig. 11) as £. famula is MACN 8493. Also figured by Ameghino as £. famula are MACN 8505, a left M1 " complete (1906, fig. 246a), and MACN 8504, a fragment of a left mandibular ramus with M2_4 complete (1906, fig. 246e, o). All of these specimens are referable to M. tehuelchum. Phonocdromus patagonicus Ameghino (1894, p. 355) was based on two speci- mens. MACN 5741, a left mandibular ramus with partial dentition is here selected lectotype. I could not locate this specimen in the MACN collection; although the box in which it was supposed to be was there, the specimen was not. Reig (1955b, p. 63) says that it is a microbiothere, however, and the measurements given by Ameghino suggest that it is referable to M. tehuelchum. The second specimen, MACN 8456, a fragment of a left mandibular ramus with alveoli of Mw and M2_4 complete, is referable to Stilotherium dissimile. This specimen is listed erroneously as 8450 by Reig (1955b, p. 64). Hadrorhynchus tortor Ameghino (1891, p. 311) was based on MACN 5738, a fragment of a right mandibular ramus with alveoli of ^-C, roots of P,_2/ P3 complete, and alveoli of Mj_2, and is listed as "tipo" in Ameghino's catalog. To Microbiotherium tortor Sinclair (1906, p. 412, pi. 62, figs. 1, 2, 2a) referred PU 15698, and Ringuelet (1953, p. 278, pi. 1, fig. 3) referred MLP 11-35. All of these specimens are referred to M. tehuelchum. Hadrorhynchus torvus Ameghino (1891, p. 311) was based on MACN 5739, a fragment of a right mandibular ramus with alveoli of C, roots of Plr P2_3 complete, and alveoli of Mx. This specimen is listed as "tipo" of the species in Ameghino's catalog and is figured as Microbiotherium torvus by Ringuelet (1953, pi. 1, fig. 5). Sinclair (1906, p. 412) regarded H. torvus as a junior synonym of M. tortor. In this study I regard both species as synonyms of M. tehuelchum. MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 35 In summary, seven of Ameghino's species are here included as junior syn- onyms in Microbiotherium tehuelchum. The points used by Ameghino to charac- terize these species are all here attributable to preservational, age and/or indi- vidual differences of the various specimens. Some linear tooth measurements of these and other specimens are given in Tables 5 and 6, and statistics for these dimensions are given in Table 7. Sample sizes for various lower tooth dimensions are quite large (N = 10-22) and CV values range between five and nine with only one value (M4 W) greater than 10. For upper tooth dimensions the sample sizes are smaller (N = 2-5) with most CV values between four and eight, and with only two values (M1 W and M4 W) greater than 10. Considering the generally large sample sizes these values are within the range expected within a population of a single species. Microbiotherium gallegosense Sinclair, 1906. Figure 11. Microbiotherium gallegosense Sinclair, 1906, p. 413, pi. 62, figs. 3, 3a; Ringuelet, 1953, p. 277. Type of M. gallegosense. — AMNH 9591, a relatively complete right mandibular ramus with alveoli of C-P2 and M3.4, and with P3-M2 complete (figured by Sinclair, 1906, pi. 62, figs. 3, 3a). Hypodigm. — Type only. Horizon and locality. — Type is from the Santa Cruz Formation, Santa Cruz Province, Patagonia, southern Argentina, and was collected by Barnum Brown in 1899 from the north bank of the Rio Gallegos near Felton's Estancia. Age. — Santacrucian. Diagnosis. — Largest known species of Microbiotherium; averages 29% larger than M. tehuelchum; lower canine large and set perpendicular in jaw; P, set at sharp oblique angle in jaw; P2 set straight in jaw; posterobasal heel on P3 large; very sharp decrease from P3 to Pl7 P3 proodont and much larger than P2; P3 same height as M^ minute foramen pierces masseteric fossa anteriorly. Measurements. — P3 L = 1.9 mm, W = l.l mm; M, L = 2.5 mm, W = 1.7 mm; M2 L = 2.6 mm, W = 1.8 mm; P,_3 L = ca 5.9 mm; M,_4 L = ca 9.5 mm; P3-M4 L = ca 11.6 mm. Comments. — This species is known only from its type and is easily distin- guished from other members of the genus by its large size. Microbiotherium gallegosense differs from the large didelphid Colhuehuapian species Pachybio- therium acclinum in being slightly smaller in size and having an absolutely and relatively larger paraconid, a smaller protoconid, and large talonid basin on M,_2. Dromiciops Thomas, 1894, p. 186 Dromiciops Thomas, 1894, p. 186. Type. — Dromiciops australis (Philippi, 1893a, p. 31). Distribution. — As for type and only known species. Diagnosis. — As for type and only known species. Dromiciops australis (Philippi, 1893a). Figures 12-17; Tables 8, 9. Didelphys australis Philippi, 1893a, p. 31; 1893b, p. 318. Dromiciops gliroides Thomas, 1894, p. 187. 36 MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 37 Type of D. australis. — The type is in the Natural History Museum in Santiago, Chile and consists of a skin and skull (inside the skin) (see Osgood, 1943, p. 49), collected near Union, Valdivia Province, Chile. Type ofD. gliroides. — British Museum (Natural History) 92.5.9.3, an aged male preserved as skin and skull, collected from Huite, near Ancud, Northeastern Chiloe Island, Chile. [This specimen was obtained by Dr. R. O. Cunningham in 1868, who referred to it (1871, p. 362) under the name Didelphys elegans.] Distribution. — Known only from south central Chile, from Concepci6n south to Chiloe Island, and east to slightly beyond the Argentine border near Lago Nahuel Huapi. They occur in dense, cool, humid forested areas in the Cordillera de la Nahuelbuta and Cordillera de los Andes, above 360 meters in elevation and below tree line (fig. 12; see Marshall, 1978a, p. 2 for a summary list of some known collection localities). Diagnosis. — Small sized microbiothere, closely resembling Microbiotherium pa- tagonicum (Santacrucian) in size and structure; body fur silky, short, dense; ears short, rounded, covered with short hair, and anterior basal prominence little developed; backs of ears covered basally with thick fur similar to that of crown, and terminally with thin yellowish hairs. Tail thick basally, tapering rapidly and evenly to tip; basal third of tail thickly clothed with shining fawn colored fur resembling that of body; terminal two- thirds of tail almost equally well clothed, but hairs straighter and nearly uni- formly dark brown; below hairs are brownish white throughout; naked part of tail confined to a narrow strip 25 to 30 mm long on underside of tip. Hands and feet are large and well developed, and limb bones are propor- tionately shorter, wider, and more robust than in species of Marmosa; digit four of hind foot is largest while two, three and five are shorter and one is shortest of all; soles of hind feet have five prominent transverse striated pads (plantar tubercules), one across base of hallux, three at bases of toes two through four, and a fifth (smallest) pad along posteroexternal border; large longitudinally striated terminal digital pads extend beyond claws. Female has a small but well-developed abdominal pouch (Mann, 1958, p. 211, figs. 3, 4) with four mammae symmetrically placed in two pairs; a medial "pseudo" vagina is well developed (Osgood, 1943, p. 49; Mann, 1958, p. 209, figs. 1, 2). Scrotum of male is shortly pedunculate and covered with cinnamon rufous hairs, identical to those covering pouch area of females (Mann, 1958, p. 209). No trace of distinct temporal ridges, supra- or postorbital processes, or a sagittal crest; auditory bullae smaller and not as elongated anteriorly as in Mi- crobiotherium tehuelchum; no trace of para- and metaconules on M1"3; canines short, upper not exceeding line drawn from tip of P3 to I1; lower canine very small, incisivo-premolariform with broad posterobasal swelling and scarcely higher than P3; P1 2 subequal in size, and smaller than P3; P,^3 increase gradually and moderately in size from P, to P3; P! not set obliquely in jaw; posterobasal heel on P2 large and well-developed; no evidence of mental foramen below M,; deep middle cleft on M3. Description. — General body color above is generally a fawn gray; the dorsal area is decidedly darker than the sides. The face is a pale gray with distinct black rings around the eyes. Crown and nape of neck are rufous brown or cinnamon, the sides of the body where the hairs have prominent subterminal white rings are paler. Large whitish patches also occur on sides of body just behind shoulder, Fig. 12. Map of south central Chile and western Argentina showing known geographic range of Drotniciops australis (after Marshall, 1978, fig. 4, and based on data in Osgood, 1943, map 3 and Gollan, 1946, pi. 3). Subspecies are: 1, D. a. australis; and 2, D. a. gliroides. 38 MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 39 Fig. 13. Dromiciops australis (Philippi, 1893a, p. 31) (Recent). AMNH 92147, skull of a female from Lota, Concepci6n Province, Chile: top, dorsal; middle, lateral; bottom, ventral views. Scale = 5 mm. (After Marshall (1978, fig. 1).] in front of shoulder, and just behind hips. The whole animal, when viewed from the side, presents an alternation of light and dark areas, with the whitish cheeks followed by a darker color below the ears, then paler on sides of the neck, and darker on the shoulder, center of the belly, and hips; each of these parts is separated from the others by paler areas. The dark of the shoulder is united above to that of the back, but that of the hips is separated by a narrow, longitudinal, light line passing approximately along the pelvic bones. These variations in color are in no case conspicuous or sharply defined. The appearance is suggestive, however, of the more distinctive color pattern of the water opos- sum Chironectes. The belly is a dirty yellowish white; the gray of the bases of the hairs shows through. The outer sides of the limbs and backs of the hands are 40 FIELDIANA: GEOLOGY Table 8. Measurements of cheek teeth of Dromiciops australis (measurements are of dentition on right side). Ml-3 P3 Ml M2 M3 M4 Ml-4 P3-M4 Specimen Sex L w L W L W L W L W L L Upper Cheek Teeth FMNH 22671 M 1.0 0.7 1.7 1.4 1.6 1.8 1.4 1.7 0.6 0.4 4.5 FMNH 22672 F 1.0 0.6 1.6 1.3 1.5 1.6 1.4 1.7 0.7 0.5 4.3 FMNH 22673 F 1.5 1.3 1.5 1.6 1.4 1.6 4.4 FMNH 22675 M 1.6 1.4 1.6 1.7 1.4 1.6 FMNH 50073 M 1.0 0.6 1.8 1.4 1.5 1.7 1.4 1.5 4.4 FMNH 50074 F 1.1 0.7 1.8 1.4 1.6 1.7 1.5 1.7 4.6 Lower Cheek Teeth FMNH 22671 M 1.1 0.6 1.5 1.0 1.5 1.1 1.4 0.9 1.0 0.6 5.5 6.5 FMNH 22672 F 1.0 0.6 1.4 0.9 1.5 1.0 1.4 0.8 1.2 0.5 5.5 6.5 FMNH 22673 F 1.4 0.9 1.5 1.0 1.4 1.0 FMNH 22675 M 1.4 0.9 1.5 1.1 1.4 1.0 FMNH 50073 M 1.0 0.5 1.5 0.9 1.5 1.0 1.3 0.8 1.0 0.6 5.3 6.2 FMNH 50074 F 1.1 0.6 1.4 0.9 1.6 1.1 1.5 1.0 1.2 0.7 5.7 6.7 Table 9. Statistics for some cheek tooth dimensions of Dromiciops australis. Dimension N OR X s cv Upper Cheek Teeth p3 L 4 1.0-1.1 1.03 0.05 4.85 W 4 0.6-0.7 0.65 0.06 9.23 M1 L 6 1.5-1.8 1.67 0.12 7.19 W 6 1.3-1.4 1.37 0.05 3.65 M2 L 6 1.5-1.6 1.55 0.05 3.23 W 6 1.6-1.8 1.68 0.08 4.76 M3 L 6 1.4-1.5 1.42 0.04 2.82 W 6 1.5-1.7 1.63 0.08 4.91 M4 L 2 0.6-0.7 0.65 0.07 10.77 W 2 0.4-0.5 0.45 0.07 15.56 M1"3 L 5 4.3-4.6 4.44 0.11 2.48 Lower Cheek Teeth P3 L 4 1.0-1.1 1.05 0.06 5.71 W 4 0.5-0.6 0.58 0.05 8.62 M, L 6 1.4-1.5 1.43 0.05 3.50 W 6 0.9-1.0 0.92 0.04 4.35 M2 L 6 1.5-1.6 1.52 0.04 2.63 W 6 1.0-1.1 1.05 0.05 4.76 M3 L 6 1.3-1.5 1.40 0.06 4.29 W 6 0.8-1.0 0.92 0.10 10.87 M4 L 4 1.0-1.2 1.10 0.12 10.91 W 4 0.5-0.7 0.60 0.08 13.33 M,_4 L 4 5.3-5.7 5.50 0.16 2.91 P3-M4 L 4 6.2-6.7 6.48 0.21 3.24 brown; the inner sides of the limbs and backs of the hind feet are dull white. The tail is moderately prehensile and its pelage is denser than in species of Marmosa (Osgood, 1943, p. 49). The presence of fur on the tail of Dromiciops (and Glironia) is perhaps correlated with a decrease in the use of the tail as a prehensile organ (Tate, 1933, p. 15). Adults caught in the Cordillera de Nahuelbuta average slightly larger in several MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 41 measurements, especially in length of head and body, than specimens of similar age taken in the Cordillera de los Andes (see table 8 in Greer, 1965). There are no significant color differences between skins from these areas (Greer, 1965, p. 103). Osgood (1943) formally recognized two subspecies, the mainland Drotniciops australis australis and the Chiloe Island D. a. gliroides. He noted (p. 50) that a specimen of D. a. gliroides was "noticeably darker and shorter-tailed than spec- imens from the mainland, indicating at least subspecific distinction. The under parts especially are darker, and the tail, which in [D. a.] australis is usually somewhat lighter below, is wholly dark." The following ranges of measurements (in mm, except weight) are from Mann (1955, p. 159, data for eight specimens), Gollan (1946, p. 192, data for six to seven specimens), and Greer (1965, p. 104, data for 23 to 26 specimens), and selected cheek tooth measurements of six specimens in the FMNH are given in table 8; all are referable to D. a. australis: weight 16.7-31.4 g; total length 195-246 mm; length of head and body 83-130 mm; length of tail 90-132 mm; length of hind foot 15-20 mm; length of forefoot 9-10 mm; height of ear at notch 17-20 mm; height of ear at crown 12-16 mm; condylobasal length of skull 23.6-27.9 mm; zygomatic breadth 14.4-16.4 mm; mastoid breadth 11.5-12.6 mm; least interorbital breadth 4.5-5.1 mm. Measurements for type of D. a. gliroides given by Thomas (1894) are: combined head and body length 126 mm; length of tail 102 mm; length of hind foot 18 mm; ear (at notch) 14 mm; basal length of skull 26 mm; greatest breadth of skull 16.8 mm; nasal length 11.1 mm; greatest nasal breadth 4.4 mm; least nasal breadth 1.9 mm; interorbital constriction 5.3 mm; length of palate 14 mm; breadth of palate outside M3 10 mm; length M1"3 4.6 mm; length Mj_4 5.8 mm. The summary of diagnostic and descriptive characters was compiled from Cabrera (1919, p. 30), Cabrera & Yepes (1960, p. 46), Fischer (1978), Mann (1955), Osgood (1943, p. 48), Thomas (1894, p. 187), Philippi (1893a,b), Walker et al. (1968, p. 17), and Wolffsohn (1913). Behavior and ecology. — Drotniciops australis is almost entirely nocturnal and in- habits dense, cool, humid forests of the Chilean Andes. Osgood (1943, p. 48) considered D. a. australis restricted to the "Valdivian Forest District" from the highest parts of the Cordillera de la Nahuelbuta east to, and beyond, the Ar- gentine-Chilean border. Moist, forested areas, at least in Malleco Province, ap- pear to be the preferred habitat, although it is not confined to this habitat, being also found in Araucaria-Nothofagus dombeyi forests, N. obliqua woods, N. obliqua- N. dombeyi forests, and in localized wooded areas having evergreens, such as Peumus boldus, Guevina avellana, Persea lingue, Myreeugenia sp., and Drimys winteri, characteristic of southern forest. Cane (Chusquea sp.) thickets in these forested areas are likely places to trap D. australis (Greer, 1965, p. 103). Mann (1955, p. 163) reported that these animals hibernate in winter when cold temperatures prevail and food is scarce. Hibernation is preceded by an accumulation of fat in the basal part of the tail (Fischer, 1978; Hunsaker, 1977). The heart rate drops from 230/minute to less than 30/minute (Mann, 1955, p. 161). Animals captured by Greer (1965, p. 105), kept alive in cages and exposed to the nighttime cold temperatures in the same region in which they were trapped, would assume a flexed position. If early morning temperatures dropped to 4.5° C or lower, the animals' bodies became stiff and cool to the touch. When placed in a warm place they soon emerged from their torpor and became quite no c* ■<* '> i— i al 42 MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 43 Fig. 15. Dromiciops australis (Philippi, 1893a, p. 31) (Recent). FMNH 22673, a left dP, (stipple) and P2 and M, (outline): a, labial; b, occlusal; c, lingual views. Scale = 3 mm. active. Their dark dense fur and small ears appear to be additional adaptations to cold environments (Mann, 1955, p. 161). Mann (1958, p. 210) reported finding nests of Dromiciops below fallen trunks, in hollow trees, and on branches. Nests in the bamboos Chusquea coleu and C. cummingi usually were about two meters above the ground. One nest (see Mann, 1958, p. 212, fig. 7), 200 mm in diameter, was spherical with an opening of 30 to 40 mm. A small wall, presumably affording some sort of protection, was found at the opening of two of the five nests observed. The nests are usually constructed of the water repellent leaves of Chusquea and are lined with grasses or moss. Four of the five nests were entirely covered with gray moss, which not only protected it against rains but closely resembled the substrate in form and color, thus concealing the nest admirably. Greer (1965, p. 103) found no nests in trees as reported by Mann, but he did find a nest in a thicket of cane (Chusquea sp.) bent to the ground by a fallen tree. One animal was trapped in front of a burrow that opened onto a dry slope next to a stump of an old roble (Nothofagus obliqua). The burrow averaged 100 mm 44 MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 45 Fig. 17. Dromiciops australis (Philippi, 1893a, p. 31) (Recent). FMNH 22673, a left dP3 (stipple), and P2 and M1 (outline): a, labial; b, occlusal; c, lingual views. Scale = 3 mm. in diameter and extended for 760 mm, sloping downward slightly before dividing into two separate smaller tunnels of undetermined length. No nests were found. Both males and females become sexually mature during their second year. They reportedly live in pairs at least during the breeding season in spring (Mann, 1958). Philippi (1893a) reported finding five young in a nest near Valdivia. Krieg (1924) observed a female with four young, but he did not specify the state of their development. Osgood (1943) found three young in the pouch of a female captured on 23 November at Peulla. In January and February 1958, Mann (1958) found two small young in an arboreal nest in the trees on the south slope of Osorno Volcano. He also reported capturing a female carrying a young male on her back. Greer (1965) trapped a female with two pouch young measuring 8.0 mm and 8.2 mm on 21 December 1960. Kirsch (in Collins, 1973, p. 60) reported having trapped a female during May, whose well-developed mammae were "[s]training against the pouch opening." From the appearance of the mammae, Kirsch surmised that the animal had large nest young at the time of capture. 46 FIELDIANA: GEOLOGY Four stages have been described in the development of the young: 1) in the pouch attached to the teats; 2) in the nest; 3) on nocturnal trips on the mother's back; and 4) in loose association with other members of the family (Mann, 1958, p. 210). In the wild, D. australis feeds largely on larvae and pupae of a variety of arthropods, especially insects, with preference for coleopterans (Mann, 1955, p. 164; Walker et al., 1968, p. 17). The intestine of Dromiciops is relatively shorter than that of Marmosa, reflecting the predominance of animal matter in the diet of Dromiciops, in contrast to the more omnivorous diet in species of Marmosa (Mann, 1955, p. 166). Greer (1965, p. 105) maintained three specimens at Michigan State University on a diet consisting of oatmeal, potatoes, powdered milk mixed with canned peach syrup, plum, pear, watermelon, applesauce, blackberries, and hamburger. The hamburger was eaten ravenously. According to Mann (1955, p. 166) Dromiciops has protection from predators in the form of strong and repugnant secretions from cutaneous glands. Greer (1965, p. 105) did not notice a disagreeable odor when the live animals were disturbed by handling as reported by Mann, nor did handling live-trapped animals provoke aggressive behavior. The voice of Dromiciops has been recorded as a chirring trill with a slight coughing sound at the end (Walker et al., 1968, p. 17). The name colocolo is but an onomatopoeia of its distinctive cry: tzchi, tzchii, kod-kod (Cabrera & Yepes, 1960, p. 46). At night, or when left unmolested for a period of time, some individuals made soft buzzing noises (Greer, 1965, p. 105). Greer (1965, p. 104) trapped Dromiciops in association with other small mam- mals, such as Oryzomys longicaudatus, Akodon longipilis, A. olivaceus, Notiomys macronyx, N. valdivianus, Phyllotis micropus, Irenomys tarsalis, and Octodon bridgesi. In the Mapuche Indian language it is called "Kongoy-kongoy;" other indig- enous names are "kunuuma" and "huenukiki" (Cabrera & Yepes, 1960, p. 46). In Chile the names "Uaca" and "monito del monte" are in general use for this animal, although both names are sometimes also applied to Marmosa elegans (Osgood, 1943, p. 50). Summary of Evolution of Microbiotheriidae Members of the South American marsupial family Microbiotheriidae are known from beds of Colhuehuapian and Santacrucian age in Patagonia, southern Argentina, in the Recent fauna from Chiloe Island and south central Chile, and from adjacent parts of Argentina. As yet, they are the only known didelphoids of Santacrucian age in all of South America. Two genera, Microbiotherium and Dromiciops, and seven species are recognized. These species are distinguished primarily on the basis of absolute size (figs. 18-21). For example, in length of M,_4 (fig. 20), all species for a known given age are readily separable one from the other. The same is true for plots of L Mj versus L M4 (fig. 21A), L P3 versus W P3 (fig. 21B), and L MA versus W Mj (fig. 21C). Relative size and minor structural differences in the canines and cheek teeth have also proven diagnostic (table 10). Microbiotherium is known from six fossil species while Dromiciops is known from one living species. Dromiciops differs from Microbiotherium as follows: 1) has a more reduced lower canine which is incisiform-premolariform in shape and A B C M4M3 M2 M1 P ' 2 <as^S0sj 3 ^Kcb^O0^ 4 c0CJ^(^<^ «3<332Si <&G>£D °^ occco Oc? ^ aS52 &£ J Fig. 18. Comparison of upper dentitions of various species of Microbiotheriidae and a caenolestid, showing relative size and proportions of teeth. All illustrations are drawn to same scale. A, labial; B, occlusal; C, lingual views. 1, Microbiotherium tehuelchum; 2, Mi- crobiotherium patagonicum; 3, Dromiciops australis; 4, Stilotherium dissimile. A 5 ^cCrC^oO^7 7 tocCcO | 8 ££r£d B M, P3 'Oj^^p^^^ CKd .aGQc^ &&£? &&&& 9C 11 c&5| 12 G ^IGESI cC? fo €3 13 14 £*b o £> cO Fig. 19. Comparison of lower dentitions of various species of Microbiotheriidae and some Didelphidae and Caenolestidae showing relative size and proportions of teeth. All illustrations are drawn to same scale. A, labial; B, occlusal; C, lingual views. 1, Micro- biotherium gallegosense; 2, Microbiotherium tehuelchum; 3, 4, Microbiotherium divisum; 5, Mi- crobiotherium praecursor; 6, Dromiciops australis; 7, Microbiotherium patagonicum; 8, Microbio- therium acicula; 9, Pachybiotherium acclinum; 10, Patene pattersoni; 11, 12, Eomicrobiotherium gaudryi; 13, Stilotherium dissimile; 14, Phonocdromus gracilis. 47 48 FIELDIANA: GEOLOGY L M1 . (in mm) 4 0 5.0 i 6.0 i 7.0 i 8.0 i 9.0 i 10.0 1 M. acicula - • M. patagonicum - • M. praecursor - ■ M. divisum - ▲ ▲ A M. tehuelchum - M. gallegosense - • • Fig. 20. Size distribution of the species of Microbiotherium as indicated by length of Mj_4. Symbols are: Colhuehuapian (triangles), Santacrucian (circles), unknown (square). is set at an anterior inclination in the jaw (in Microbiotherium the canine is larger and is set nearly perpendicular in the jaw); 2) Pj is set straight in jaw relative to rest of tooth row (not obliquely as in Microbiotherium); 3) posterobasal heel on P2 is large and well-developed (it is small but distinct in Microbiotherium); 4) a small mental foramen below M, is absent (one is present in Microbiotherium); 5) auditory bulla is relatively smaller and not as elongated anteriorly as in Micro- biotherium; 6) a supraorbital process is absent (a large one is present in Micro- biotherium); 7) para- and metaconules are absent on M1"3 (small ones are present in Microbiotherium); 8) middle cleft on M3 is deep (not shallow or absent as in Microbiotherium); and 9) M4 is smaller relative to M3 than in species of Micro- biotherium. Microbiotherium acicula is the smallest and most generalized of known micro- biotheres. In comparable linear tooth dimensions it averages 15% (n = 10, OR 0-30%) smaller than the contemporaneous Santacrucian species M. patagonicum. Pu however, averages slightly larger and M4 is smaller relative to M3 — with a smaller talonid in M. acicula than in M. patagonicum- — but in morphological fea- tures these species are similar point for point. In other respects M. acicula serves as a proto- or morphotype from which may be derived all other known species of Microbiotheriidae. Microbiotherium patagonicum is intermediate in size between the contempora- neous, smaller M. acicula and the contemporaneous, larger M. tehuelchum, and differs in average size from each by about 15% in linear tooth dimensions. M. patagonicum differs structurally from M. tehuelchum as follows: an absolutely and relatively smaller lower canine; a smaller posterobasal heel on P3; but a slight size decrease in P3 to Pj (not a sharp one as in M. tehuelchum); P3 less than height of My protocone onM1"3 narrower anteroposteriorly; labial ridge of stylar shelf on M1"3 less distinct; a very shallow middle cleft on M2 (absent in M. tehuelchum); and M4 smaller relative to M3 than in M. tehuelchum. Microbiotherium patagonicum is extremely similar in size and molar morphology -C ? u 0) sg 3 -C jt ■ 5 c IT) u D - P 5 r- C O "So o a. VI C o •X! T i £? CO « 1 p E _ E CO a. o C 5 ~ J^ 3 S E tS tj j= a* 5 ■g 1 Ti •a ^ c rj <a C 6 ft u. Z 49 iff II, <U.S 60 T3 ra C C ^ s s -a o S I* ai B" > 60 01 > 3 •ST 3 5» H ra 1-3 !| Jrf 35 T3 .. ra ■a * -a So z 6 5 £ ■■P s ss S § -2 ■I 6 1 B c (0 0> bS <— *• h — -. B P 5 b Xi m m O P 5 J2 « S3 60 a. 8 5 •JrQ ■2 «2 ^r •« « € s c s ^ o> o ex .y en C % 0 c £ xl £ E •2 § I & o 3 § 3 pa a y 60 01 £ .3" >h rv ra x ra •* — i o ~Z £? c 8T!- £ •~ ra »-. 60 6 gl .2-3 ^3 m m 60 c _, Ji 3 X — > en £ ra - £> E 2 5 fl S S g 5 p B..S 5 > Oi C u ■■5 -9 fl S "5 "S (fl 01 . — . ni tn — ~ 60 c _ra 3 3 o U u T3 X! -a C 60 c 01 — nj a E a. .5, in — - 60 ra c — ra S n o ra ~ o - is JH 3 - a Of _c ^ Z -C "« Jl °^ « ° o o « « o tn •43 o; O ra ■s 01 7^ h e^ ? B Z z£ ■O £ -a P ~ C t, 6 « ij w 5 » E 2 ai a. «J § .2 60 O g « 13 ■J3I (A SIB o -2 60 O) »-i ^60 JS C ^ * n 2 g tn C 9 & c 3 ■J3 C <« u <" § -2 S (ft "U •9 ^ ii -a S £ oTO -43 S «S cTcK" T3 C-i 01 p Kl ** >? He. So. 01 _ -Ss * e C 2 ra o> 5 o « •c vc o O *£ '*> tn ra > a, ■& «i oi o g 5 vk ra ra 60 O. S £ o C H C § goof 2 2 S 7 6 Q £? 3-SP-S .2 o * o ^ § ' JS ^^ 2 5 2 5 % ■ 1 S E ^ 2 S °E 5 X> jj o — 6 S c 8 ^ S SoeCS I ^ « S c y "3 ° | 2x « E <2 >*- 7 ra .2 60 E 'S 2 •^ >° Ol „_ ra _, C 111 50 1 8 I * Jr •! •&5> 3 c c £ C £ 5K a — 5! * =5 II ■5 £ 60 O ^ <0 41 ^ o __ o 0 O.J) ♦; .? IS J T3 •" t; oi ^s « — D ra y tn ^ 60 •c.s i 2 g. C u C •j- -i - 2 B.„ S £ § I < fe •**** '5 ! •a "5 2 S 9J i * .„ 01 -C 3 60 <" C .fi en u S IP 3 M ? ° 8. at s i e *.§ *« c V o 60 So •2 -~ * S c <n © a I8 II3 ~ 60 <a u « » •£ ci- _ c n TJ -C JZ 01 W __ i~ 1 ^ 3 .£> O^ o 0) =3 01 01 * * | "3-0 3! 2 ctjxi j§ «j oi « o 01 9- O < (ft "** 5 P-e ■r. C 01 — . o — s i * (X c § 5 C J= E V c 01 3 fl ■" 3 & * £ "to o H .5 n >^-K •SP-e a £ "O u c Oi T* T3 *- -c en — 5 0 c 1 £ 3 5 2? "(3 .£ 3 (A | 1 o S ?c o — c ■£ S S O 3 O -C * fc is 2; a» s s .£?■£ 3 I - 01 3 .ga{ c S ^ :~ C oi a — ^Oi c o 8 5 J 6 B. § 2 c *- o SO II i •s^n «. Ill ■a 2s <£3£ c c 01 o> — ^^ ° •a -o oi S S in c c iS « o B Si o c Oi i'2? 51 52 FIELDIANA: GEOLOGY to the living species Dromiciops australis. In the known comparable characters these species differ as follows: D. australis 1) has a more. reduced lower canine which is incisiform-premolariform and set with an anterior inclination in the jaw (the canine is larger and set more perpendicular in M. patagonkum); 2) P, set straight relative to rest of tooth row and not at slight oblique angle as in M. patagonkum; 3) has a large posterobasal heel on P3; 4) shows a moderate, not slight, size decrease from P3 to P,; 5) lacks a small mental foramen below M^ 6) lacks para- and metaconules on M1"3; 7) has deep, not shallow, middle cleft on M3; and 8) has a smaller M4 relative to M3 than in M. patagonkum. These differ- ences are minor and I here recognize M. patagonkum as the probable Santacrucian ancestor of D. australis. The major changes involved in such a lineage involve either loss or size reduction in some features (e.g., mental foramen below M,, canine, conules, M4) or relative and absolute size increase in others (e.g., pos- terobasal heel on P3). Mkrobiotherium praecursor is known only from its type, which is of uncertain age and provenance. It is of medium size and is indistinguishable in size and structure from smaller specimens of the Santacrucian species M. tehuelchum, and is slightly smaller than known specimens of the Colhuehuapian species M. divisum. Because of the uncertain age and locality data, M. praecursor is tentatively retained here as a valid species. Mkrobiotherium divisum is the only known microbiothere of Colhuehuapian age. It is of medium size and is similar in size and structure to larger specimens of the Santacrucian species M. tehuelchum, and differs from the latter as follows: 1) Pj is set at only a slight oblique angle in jaw; 2) posterobasal heel on P3 is distinct but not large; and 3) a more gradual size decrease occurs from P3 to P^ These differences are minor and I have recognized M. divisum as the probable Colhuehuapian ancestor of M. tehuelchum. In the evolution of this lineage there occurred a slight reduction in size of the premolar region of the jaw, resulting in a crowding of the premolars such that 1) P, is set at a sharp oblique angle in the jaw and 2) a sharp decrease occurs in size from P3 to P^ The Santacrucian species Mkrobiotherium tehuelchum is the most abundant microbiothere known in the fossil record. It is of medium size and averages 16% larger in linear tooth dimensions than the contemporaneous M. patagonkum and 29% smaller than the contemporaneous M. gallegosense. M. tehuelchum differs from all other microbiotheres where comparable features are preserved, as fol- lows: 1) P§ (in some specimens) set at slight oblique angle in jaw; 2) P1 single rooted (or nearly so) with no trace of a posterobasal heel; 3) protocone on M1"3 broader anteroposteriorly and more robust, and labial ridge on stylar shelf better developed; and 4) a middle cleft on M2 absent. As noted above, M. tehuelchum is here regarded as the direct Santacrucian descendant of M. divisum. The largest known species of microbiothere is M. gallegosense which is known only from its type, collected from beds of Santacrucian age of southernmost Patagonia. Apart from averaging 29% larger than the contemporaneous M. te- huelchum, it differs from that and other microbiotheres in having a minute fora- men which pierces the masseteric fossa anteriorly. The only living microbiothere is Dromiciops australis. This small microbiothere is restricted to a Chiloe island and a small area of south central Chile and adjacent Argentina. It is, as noted above, regarded as the living descendant of the San- tacrucian species Mkrobiotherium patagonkum. These species are extremely similar k a. 53 54 FIELDIANA: GEOLOGY in their dental structure, suggesting that this lineage has been little changed during the last 20 million years of its evolutionary history. The seven recognized species of microbiothere are listed in Figure 22 in the order of their chronostratigraphic occurrence and with indication of their prob- able phylogenetic relationship. The proposed phylogeny is based largely on the data set summarized in Table 10, although it includes consideration of other features discussed in the text. REVIEW OF SUPPOSED MICROBIOTHERES In the following section the taxonomic history of genera and species 1) formally assigned at one time or another to the Microbiotheriidae; 2) implicated in the ancestry of this family; or, 3) in which specimens of microbiotheres were included erroneously as syntypes in the original descriptions, are discussed and the sys- tematic problems clarified. For all of these taxa an attempt is made to give complete listings of generic and specific synonymies, complete literature cita- tions, and hypodigms which include all known specimens. In cases where mi- crobiotheres were among syntypes, as in the caenolestid species Stilotherium dissimile and Phonocdromus gracilis, the taxonomic history of each of the disputed specimens is discussed in the Systematic Section of the relevant microbiothere species (see above). Superfamily INDETERMINATE Ideodelphys Ameghino, 1902d, p. 43 Ideodelphys Ameghino, 1902d, p. 43. Ideodidelphys Schlosser, 1923, p. 442. Type. — Ideodelphys microscopicus Ameghino, 1902d, p. 43. Distribution. — As for type and only known species. Diagnosis. — As for type and only known species. Ideodelphys microscopicus Ameghino, 1902d. Figure 23. Ideodelphys microscopicus Ameghino, 1902d, p. 43; Simpson, 1938, p. 3; 1948, p. 34 (nomen vanum); Ringuelet, 1953, p. 287. Ideodelphys microscopica Reig, 1955a, p. 122 (nomen vanum). Type. — MACN 10344, an edentulous right mandibular ramus with eight com- plete alveoli. Hypodigm. — Type and possibly AMNH 28423, an edentulous mandibular ra- mus with eight complete alveoli. Horizon and locality. — Type is from the barranca south of Lago Colhue-Huapi and was collected by Carlos Ameghino; AMNH 28423 is from Canadon Vaca and was collected by members of the Scarritt Expedition to Patagonia in 1930; both localities are in Chubut Province, Patagonia, southern Argentina. Age . — Ca samay oran . Diagnosis. — Indeterminate. Comments. — The eight complete alveoli in the type measure 5.4 mm in length and those in AMNH 28423 are 5.8 mm. It appears that the alveoli in the type represent P2-M2 and in the referred specimen they are P3-M3, although this is MARSHALL: SYSTEMATICS OF MICROBIOTHER1IDAE 55 Fig. 23. Ideodelphys microscopicus Ameghino, 1902d, p. 43 (Casamayoran). MACN 10344 (type), an edentulous right mandibular ramus with eight complete alveoli: a, labial; b, occlusal; c, lingual views. Scale = 5 mm. not certain. In both, the alveoli increase gradually in size from front to back. In AMNH 28423 a large mental foramen occurs below the 5th complete alveolus. Ideodelphys microscopicus thus is virtually indeterminate and the genus and species are, for all practical purposes, nomina vana (Simpson, 1948, p. 34; Reig, 1955a, p. 122). Ringuelet (1953, p. 287) provisionally included this species in the subfamily Microbiotheriinae. Alternatively, the specimens assigned to this spe- cies possibly could be referable to a species of Eomicrobiotherium (see below), but this too is only speculative. Superfamily DIDELPHOIDEA (Gray, 1821, p. 308) Osborn, 1910, p. 515 Family DIDELPHIDAE (s.l.) Gray, 1821, p. 308 Coona6 Simpson, 1938, p. 1 Coona6 Simpson, 1938, p. 1. Type. — Coona pattersoni Simpson, 1938, p. 2. Distribution. — As for type and only known species. Diagnosis. — As for type and only known species. Coona pattersoni Simpson, 1938. Figure 24. Coona pattersoni Simpson, 1938, p. 2, fig. 1; 1948, p. 34, fig. 3; Ringuelet, 1953, p. 301. 6 In accordance with Articles 27 and 32c of the International Code of Zoological No- menclature (Stoll et al., 1961, 1964) the diacritic mark is dropped from the name originally spelled Coona. > . « s ■S 6 2 II 2 8 a I 03 03 00 o x if < ,- 1? c 23 °i 6-2 3.1 -I d° . c oo > 2 ° 9 -r a ? .§1 en £ "5 ii & &- II P 3 -^ S «| a I 56 MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 57 Type.— AMNH 28907, a left mandibular ramus with alveoli of P3-M2/ M3 com- plete, and M4 with talonid complete and with crown of trigonid broken (figured by Simpson, 1938, fig. 1; 1948, fig. 3). Hypodigm. — Type only. Horizon and locality. — Casamayor beds, green bentonite series ("bird clay") at Canad6n Hondo, Chubut Province, Patagonia, southern Argentina. Collected by C. S. Williams and G. G. Simpson as members of the Scarritt Expedition to Patagonia, March 1931 (locality is shown as "E" in fig. 1 in Schaeffer, 1947, p. 3). Age. — Casamayoran . Diagnosis. — Didelphoids of small size; P3 (alveoli) about as large as M,; M3 (and probably also M,-^ with talonid markedly wider than trigonid, paraconid slightly higher than entoconid, metaconid intermediate between paraconid and proto- conid, trigonid moderately elevated and not compressed, hypoconulid very small, nearly internal, and near entoconid; M4 talonid as wide as trigonid or slightly wider and of about same area, fully basined and similar to that of M3 with three distinct cusps, hypoconulid not enlarged and similar to that in M^ Mi well developed, but smaller than M3 (after Simpson, 1938, p. 1). Measurements. — M3 L = 2.6 mm, W=1.8 mm; M4 talonid W = 1.4 mm; M,_4 L = ca 10.0 mm. Comments. — Simpson (1938; 1948, p. 36) emphatically stated that the most important characteristic of Coona pattersoni is that it is definitely not a micro- biothere. However, he did entertain the possibility that it was an ancestral mi- crobiothere. In defense of this view he noted that the M4 in microbiotheres is very reduced in size with a narrow, nearly unbasined heel with one distinct posteromedian cusp. Coona on the other hand has a large talonid, and the M4 is only slightly smaller than the M3. In addition, he noted that microbiotheres typically have the hypoconulid of Mj_3 more median in position than in Coona and more co- ordinate with the hypoconid and entoconid as an element of the talonid rim. I concur with Simpson's views and support the nonmicrobiothere position of Coona. Simpson (1964, p. 7) questionably referred a second species to Coona, ?C. gaudryi. This species is here made the type of a new genus (see p. 58) and is shown to share features with typical microbiotheres of the genus Microbiotherium, although it is not itself regarded a microbiothere. Eomicrobiotherium nov. gen. Etymology. — Eos Gr. dawn, morning, early, east; micro- Gr. small, little; bios Gr. life; ther Gr. a wild beast; -ium Gr. suffix added to noun and verb stems denoting groups. Name given in reference of this Eocene age taxon and the similarity of its members with species of the mid-Tertiary genus Microbiotherium. Type. — Eomicrobiotherium gaudryi (Simpson, 1964, p. 7). Distribution. — Casamayor beds, Patagonia, southern Argentina. Diagnosis. — Small size; P3 about subequal in height to but slightly shorter than M„ and with a large and distinct posterobasal heel; M, slightly shorter than M2_3 which are subequal in length; M,_3 subequal in breadth, and with talonids broader than trigonids; M4 smaller than M3 but not to degree that occurs in species of Microbiotherium; protoconid much larger than para- and metaconids 58 FIELDIANA: GEOLOGY on Mj_3; paraconid much smaller than metaconid on M„ relatively larger on M2, and subequal in size to metaconid on M3; metaconid set slightly posteriad of protoconid; anterobasal cingulum absent on M^, and only weakly developed on M3. Eomicrobiotherium gaudryi (Simpson, 1964). Figures 25-26; Table 11. ?Coona gaudryi Simpson, 1964, p. 7, fig. 1; 1967, p. 7, fig. 1. Type. — MNHN Tournouer Collection No. 1, two fragments of mandibular rami, probably of same individual. No. la, a fragment of a right mandibular ramus with roots of P3, and M^ complete; and No. lb, a fragment of a left mandibular ramus with roots of Pt.2 and M^, and with P3 and M3 complete. If these fragments prove to be of different individuals, the right side is to be taken as lectotype (Simpson, 1967, p. 7). The teeth of both specimens are figured by Simpson (1964, fig. 1; 1967, fig. 1). Hypodigm. — Types only. Horizon and locality. — Casamayor beds at the Barranca south of Lago Colhue- Huapi, Chubut Province, Patagonia, southern Argentina. Collected by A. Tour- nouer from his "Coli-Huapi" locality (see Tournouer, 1903, p. 472). Age. — Casamayoran. Diagnosis. — Molars relatively and absolutely longer, and trigonids relatively and absolutely lower than in E. gutierrezi. Comments. — Simpson (1964, p. 7; 1967, p. 7) questionably assigned this species to the genus Coona, noting (1967, pp. 7-8) that, in both, the talonid is distinctly wider than the trigonid, the trigonid is moderately elevated and not definitely compressed, the paraconid is slightly higher than the entoconid, the metaconid is between the paraconid and protoconid in height, and the hypocon- ulid is small and near the entoconid. Fig. 25. Eomicrobiotherium gaudryi (Simpson, 1964, p. 7) (Casamayoran). MNHN Tournouer Collection no. la (type), a fragment of a right mandibular ramus with M:_2 complete: a, labial; b, occlusal; c, lingual views. Scale = 2 mm. MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 59 Fig. 26. Eomicrobiotherium gaudryi (Simpson, 1964, p. 7) (Casamayoran). MNHN Tour- nouer Collection no. lb (type), a fragment of a left mandibular ramus with roots of P|_2 and M,^, and with P3 and M3 complete: a, labial; b, occlusal; c, lingual views. Scale = 4 mm. However, there are important features in which these species differ and for which I feel they warrant generic separation. The species gaudryi differs from pattersoni in that it is smaller in size (it averages 45% smaller in comparable linear tooth dimensions), lacks a prominent anterobasal cingulum on the M3/ in the M3 para- and metaconid being about subequal in size (in pattersoni the paraconid is much smaller than the metaconid), and has metaconid set posteriad of the protoconid in gaudryi but directly linguad of the protoconid in pattersoni. On the basis of these features I place gaudryi along with the species gutierrezi del Corro (1977) (see below) in a new genus, Eomicrobiotherium. Of the mid-Tertiary members of the genus Microbiotherium, E. gaudryi compares more closely with M. patagonicum: i.e., both are of similar size and lack a distinct anterobasal cingulum on M,_2 but have a small one on M3. These species differ in that E. gaudryi has a much narrower M, trigonid and a more reduced paraconid; the metaconid is set posteriad of the protoconid (not labiad as in species of Microbiotherium); and the talonid basin is relatively smaller. The former two features are here regarded as derived, and their presence in £. gaudryi suggests that the species is too specialized to be ancestral to any later member of the Table 11. Measurements of lower cheek teeth of Eomicrobiotherium gaudryi. Specimen MNHN (TC #la) MNHN (TC #lb) Pi M, M, M3 1.3 W 0.8 L 1.4 W 1.0 L 1.6 W 1.0 w 1.6 0.9 60 FIELDIANA: GEOLOGY genus Microbiotherium. For this reason, E. gaudryi (and E. gutierrezi) are not included in the Microbiotheriidae. However, E. gaudryi shares these same derived features with the large Col- huehuapian species Pachybiotherium acclinum. In P. acclinum these features are more pronounced (i.e., there is a greater reduction of the paraconid; more pos- teriad position of metaconid relative to protoconid). It is conceivable to envision an ancestral-descendant relationship for these species if it is granted that in the evolution of this lineage there occurred an elaboration of these features, along with size increase and a reduction in breadth of the talonid and increase in importance of the trigonid. This view is followed here. Eomicrobiotherium gutierrezi (del Corro, 1977). Figure 27. Microbiotherium gutierrezi del Corro, 1977, p. 32, fig. on p. 32. Type. — MACN 18288, a fragment of a left mandibular ramus with roots of P,, P3, M2, M4; and with P2, Mu and M3 present, latter is missing para- and metaconid (figured by del Corro, 1977, fig. on p. 32). Hypodigm. — Type only. Horizon and locality. — Collected from the green bentonite series ("bird clay") at Canadon Hondo, Chubut Province, Patagonia, southern Argentina, by Or- lando Gutierrez in February 1975 (locality is shown as "E" in fig. 1 in Schaeffer, 1947, p. 3). Age . — Casamay oran . Diagnosis. — Molars relatively and absolutely shorter, and trigonids relatively and absolutely lower than in E. gaudryi. Comments. — I attempted to study the type during a visit to the MACN in November 1979, although the specimen could not be located and its whereabouts is still unknown. Comparison of this species with other taxa thus is based solely on del Corro's (1977) comments and on the two line drawings of the type which he included as figures on page 32 of his publication. Del Corro (1977, p. 32) described the type as follows: El alveolo del tercer premolar tapado aunque por lo que se distingue es algo mas grande que el anterior, el alveolo del cuarto molar tambien esta tapado, aparentemente Fig. 27. Eomicrobiotherium gutierrezi (del Corro, 1977, p. 32) (Casamayoran). MACN 18288 (type), a fragment of a left mandibular ramus with roots of P,, P3, M2/ M4; and with P2/ M„ and M3 present (redrafted after del Corro, 1977, fig. on p. 32): a, labial; b, occlusal views. Scale = about 5 mm. MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 61 casi del mismo tamano que el molar anterior aunque, ligeramente mas angosto hacia el extremo posterior. El area del primer molar correspondiente al trig6nido y tal6nido casi iguales, tanto en largo como en ancho, con una profunda cubeta. El segundo premolar, ligeramente alabeado de adelante hacia atras, un poco despu£s del apice y angosto. Los molares con la corona alta. No se observa cingulo. Del Corro (p. 32) also gave the length and breadth of "molar 2°" as 1.0 and 0.99 mm, respectively. However, as there is no M2 preserved in this specimen it is not clear to which tooth these measurements belong. Nevertheless, the measurements and figures given by del Corro do indicate that the molars in Eomicrobiotherium gutierrezi are relatively shorter than but are of approximately the same breadth as those in E. gaudryi in which the M, L = 1.4 mm and W = 1.0 mm, and M3 L = 1.6 mm and W = 0.9 mm (see table 11). Fur- thermore, in £. gaudryi there is no anterior cingulum on the M, as in £. gutierrezi, but there is a small cingulum on the M3 in E. gaudryi, a feature which del Corro did not observe in £. gutierrezi. However, the anterior part of the M3 is broken in £. gutierrezi (see fig. 47) and the absence of a cingulum in this area may be due to this breakage. Use of this feature as a specific character to separate these species must thus remain tentative. The trigonid is only slightly higher than the talonid on both M, and M3 in E. gutierrezi, while in E. gaudryi the trigonid is much higher than the talonid (compare figs. 43 and 45 with 47). These features may represent specific differences, but they also could be attributable to more advanced occlusal wear in E. gutierrezi, a feature which I observed when the specimen was collected but which del Corro did not mention in the description of the species. Tentatively, these observed differences, for want of others, are here used to distinguish E. gaudryi from E. gutierrezi. It is cautioned, however, that direct comparison of the types of these species may prove them to be synonymous; their specific separation has yet to be adequately demonstrated. Del Corro compared his new species gutierrezi only with the Santacrucian species Microbiotherium conspicuus (Ameghino, 1891) as that species was defined by Ringuelet (1953). He noted that these species were very similar and accord- ingly assigned gutierrezi to the genus Microbiotherium. The species gutierrezi dif- fers from species of Microbiotherium in the same features as does Eomicrobio- therium gaudryi (see above), and for this reason I here refer gutierrezi to the genus Eomicrobiotherium. E. gutierrezi, as is E. gaudryi, is here regarded as too specialized to be ancestral to any later known microbiothere. Pachybiotherium Ameghino, 1902e, p. 123 Pachybiotherium Ameghino, 1902e, p. 123; Simpson, 1932b, p. 3 (as probable synonym of Microbiotherium); Ringuelet, 1953, p. 277 (as synonym of Microbiotherium). Type. — Pachybiotherium acclinum Ameghino, 1902e, p. 123. Distribution. — As for type and only known species. Diagnosis. — As for type and only known species. Pachybiotherium acclinum Ameghino, 1902e. Figure 28; Table 12. Pachybiotherium acclinum Ameghino, 1902e, p. 123; Reig, 1955a, p. 121. Microbiotherium acclinum Ringuelet, 1953, p. 277. Type. — MACN 52-370a, a fragment of a left mandibular ramus with alveolus of C; roots of P,, and M2; anterior root and portion of posterior alveolus of M4; and P2-M, and M3 complete. 62 MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 63 Table 12. Measurements of lower cheek teeth of Pachybiotherium acclinum. P2 P3 M, M2 M3 P,_3 MI3 Specimen LWLWLW L LW L L MACN 52-370a 2.0 1.2 2.9 1.7 3.3 2.0 ca 3.4 3.1 2.0 ca 7.0 9.7 MACN52-370b 2.8 1.7 ca 8.0 Hypodigm. — Type and MACN 52-370b, a fragment of a right mandibular ramus with alveoli of C-P„ roots of P2/ P3 complete and anterior alveolus of M,. Horizon and locality. — Both specimens were collected by Carlos Ameghino from Colhu£-Huapi beds, probably from the Barranca south of Lago Colhu£-Huapi, Chubut Province, southern Argentina. Age. — Colhuehuapian. Diagnosis. — Largest known species of Colhuehuapian didelphoid; mandibular ramus very deep throughout its length and bowed or curved labially; mandibular symphysis extends posteriorly to point below anterior root of P,; a large mental foramen occurs below P,; C, judging from its alveolus, was very large and set perpendicular to jaw; P^ are much smaller than P3; P„ judging from its roots, was slightly smaller than P2 and was set at slight angle in jaw relative to rest of tooth row with anterior root located labiad of posterior root; P2 is broad for its length and is bulbous basally, with a small posterobasal heel and a small pos- terobasal cuspule; P3 with large pointed cusp anteriorly and low, broad heel with a posterolingual cusp posteriorly; M, is equal in breadth but is slightly longer than M3; trigonid and talonid of M! are of about equal width; trigonid of M3 is slightly broader than talonid; paraconid is very reduced on M, and M3 (more so on M,); metaconid well-developed but relatively smaller, compared with pro- toconid on M„ than M3; talonid basin is broad with a very shallow basin; hy- poconid and entoconid are subequal in size on M, and M3; hypoconulid is about half size of entoconid on M, and separated from it by a distinct space, while on M3 these cusps are subequal in size and united basally; M4, judging from the size of its anterior root and part of the posterior alveolus, was large and ap- parently subequal to M3 in size. Comments. — Ameghino (1902e, p. 123) originally placed Pachybiotherium ac- clinum in the family Microbiotheriidae, within his group Pedimana. Simpson (1932b, p. 3) suggested that Pachybiotherium "is probably synonymous with Mi- crobiotherium" and formally recognized this synonymy in his classification of 1945 (Simpson, 1945, p. 42). This synonymy was followed by Ringuelet (1953, p. 277), who concluded . . . estoy de acuerdo con Simpson en considerar que el hecho de poseer la rama mandibular arqueada, y siendo unico caracter distinctivo, no tiene valor suficiente como para constituir con £1 un genero distinto. Reig (1955a, p. 121), based on an unpublished collaborative study with Bryan Patterson, recognized Pachybiotherium as distinct from Microbiotherium and reaf- firmed its microbiothere affinities. Recently, Reig (in press) proposed that Pach- ybiotherium is not a microbiothere but is the earliest known member of the didelphid subfamily Caluromyinae. It is my opinion that Pachybiotherium acclinum is not a microbiothere (s.s.), but that it did parallel the microbiotheres in many of its specializations (e.g., bowed mandibular ramus; large canine oriented perpendicular in jaw; loss of anterior basal cingula on molars; P3 proodont with large posterobasal heel; reduction in 64 FIELDIANA: GEOLOGY size of paraconid). P. acclinum is not a microbiothere in the strict sense because it has a large M4 (judging from its anterior root); the talonid on M3 is narrower (not wider) than the trigonid; and the paraconid is much more reduced, espe- cially on the Mt. In addition, P. acclinum appears too specialized in the same feature (as well as in the size and structure of the canine, great reduction in size of P, and P2, and posterior position of the metaconid) to be ancestral to any known living Caluromyinae. For these reasons I recognize P. acclinum and the Casamayoran species Eomicrobiotherium gaudryi and E. gutierrezi as members of the family Didelphidae (s.l.) which parallel microbiotheres in aspects of dental and mandibular evolution. Superfamily CAENOLESTOIDEA (Trouessart, 1898, p. 1205) Osborn, 1910, p. 517 Family CAENOLESTIDAE Trouessart, 1898, p. 1205 Stilotherium Ameghino, 1887 Stilotherium Ameghino, 1887, p. 7. Garzonia Ameghino, 1891, p. 307. Parhalmarhiphus Ameghino, 1894, p. 356. Halmarhiphus Ameghino, 1891, p. 308. Type of Stilotherium. — S. dissimile Ameghino, 1887, p. 7. Type of Garzonia. — G. typica Ameghino, 1891, p. 307. Type of Parhalmarhiphus. — P. annectens (Ameghino, 1891, p. 307). Type of Halmarhiphus.7 — H. nanus Ameghino, 1891, p. 308. Diagnosis. — As for type and only known species. Known range. — As for type and only known species. Stilotherium dissimile Ameghino, 1887. Figures 29-31; tables 13, 14. Stilotherium dissimile Ameghino, 1887, p. 7; 1889, p. 265; 1894, p. 358; 1897, p. 499, fig. 75; 1903, pp. 115, 167, figs. 33, 92; 1904b, p. 45, fig. 32; 1906, p. 357, fig. 196; Schlosser, 1923, p. 27, fig. 390; Rusconi, 1933, p. 248, fig. 8; Reig, 1955b, p. 62, fig. 1; Marshall, 1980, p. 35, figs. 4-6. Pichipilus exilis Ameghino, 1891, p. 307; 1894, p. 351. Garzonia captiva Ameghino, 1891, p. 308; 1894, p. 355. Garzonia minima Ameghino, 1891, p. 308; 1894, p. 355; 1903, pp. 157, 186, figs. 81, 121; 1904b, p. 45, fig. 31; 1906, p. 357, fig. 197. Halmarhiphus didelpoides (sic) (partim, not lectotype), Ameghino, 1891, p. 308. Parhalmarhiphus didelphoides (partim) Ameghino, 1903, pp. 157, 163, fig. 86. Halmarhiphus didelphoides (partim) Ameghino, 1894, p. 357; 1899, fig. 4; 1903, p. 157, fig. 80; 1906, p. 357, fig. 195; Simpson, 1930, p. 56. Garzonia typica Ameghino, 1891, p. 307; 1893, p. 80, fig. 5; 1894, p. 354, fig. 41; 1898, p. 186, fig. 51; 1903, p. 168, fig. 93; Schlosser, 1925, p. 27, fig. 39A. Garzonia tipica (sic) Ameghino, 1903, p. 167, fig. 91. Stilotherium typica Reig, 1955b, p. 62. Stilotherium typicum Reig, 1955b, p. 63; Pascual & Odreman Rivas, 1971, p. 392. Garzonia annectens (partim) Ameghino, 1891, p. 307. Parhalmarhiphus annectens (partim) Ameghino, 1894, p. 357. Halmarhiphus nanus (partim) Ameghino, 1891, p. 308; 1894, p. 357; Sinclair, 1906, p. 421, fig. 7, pi. 63, figs. 9, 9a. 7 "H. didelphoides has been considered the type of this genus, but as Ameghino later suggested its removal to a distinct genus he cannot so have considered it" (Simpson, 1930, p. 56). MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 65 Phonocdromus patagonicus Ameghino, 1894, p. 355 (partim, not lectotype); Reig, 1955b, p. 64 (invalid). Garzonia patagonica Sinclair, 1906, p. 424, pi. 63, figs. 8, 8a, 10-13. Neotype of S. dissimile. — MACN 8464, a fragment of a right mandibular ramus with complete dentition, missing only tip of large incisor (figured by Reig, 1955b, fig. 1; Pascual & Herrera, 1975, pi. 1, fig. 2; Marshall, 1980, fig. 6) (selected neotype by Reig, 1955b, p. 62). Type of P. exilis. — MACN 5698, a fragment of a left mandibular ramus with alveoli of M2, M3 present but very worn, and alveoli of M4 (listed as "tipo" in Ameghino's catalog). Type of G. captiva. — MACN 5706, a right mandibular ramus with P3-M2 com- plete, and alveoli of M3_4 (listed as "tipo" in Ameghino's catalog). Type ofG. minima. — MACN 5709, a fragment of a right mandibular ramus with alveoli of M„ and M2_4 complete (listed as "tipo" in Ameghino's catalog). Type of G. typica. — MACN 5700, a fragment of a right mandibular ramus with Fig. 29. Stilotherium dissimile Ameghino, 1887, p. 7 (Santacrucian). MACN 8427, a frag- ment of a rostrum with left C, P2 and P3 (same individual as in fig. 30): a, labial; b, occlusal; c, lingual views. Scale = 5 mm. 66 FIELDIANA: GEOLOGY Fig. 30. Stilotherium dissimile Ameghino, 1887, p. 7 (Santacrucian). MACN 8426, a frag- ment of a right maxillary with M2"3; and MACN 8431, an isolated right M1 (same individual as in fig. 29): a, labial; b, occlusal; c, lingual views. Scale = 5 mm. Abbreviations are: pr, protocone; pa, paracone; me, metacone; hy, hypocone; ic, intermediate conule. P2-3 and M]_2 complete (listed as "tipo" in Ameghino's catalog; posterior part of ramus with M3_4 has been lost since specimen was figured by Ameghino, 1893, fig. 5; 1894, fig. 41; 1898, fig. 51). Type of G. annectens. — MACN 5703, a fragment of a left mandibular ramus with M,_4 complete but worn (listed as "tipo" in Ameghino's catalog). Type of H. nanus. — MACN 5720, a fragment of a right mandibular ramus with alveoli of M, and M4, and M2_3 complete but worn (listed as "tipo" in Ameghino's catalog). Hypodigm. — The seven types and MACN 5701, a fragment of a left mandibular ramus with P2 complete, alveoli of P3 and anterior root of Mu talonid of Mu M2 and trigonid of M3 complete but very worn; MACN 5702, a fragment of a right mandibular ramus with Pj_2 complete, roots of P3, M,_2 complete, and alveoli of M3_4; MACN 5708, a fragment of a right mandibular ramus with alveoli of Mj_2 and M4, and M3 complete (labeled G. captiva); MACN 5710, a fragment of a right mandibular ramus with Mj _3 complete, and roots of M4 [labeled G. minima. The jaw figured by Ameghino (1903, figs. 81, 121; 1904b, fig. 31; 1906, fig. 197) as G. minima is apparently based, at least in part, on this specimen. The part of the jaw from M, back is probably based on MACN 5710, as this specimen lacks MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 67 Table 13. Measurements of cheek teeth of Stilotherium dissimile. Ml M2 M3 M4 Ml-4 Specimen L W L W L W L W L Upper Cheek Teeth MACN8426 1.6 1.8 1.0 1.4 MACN 8431 1.9 1.8 Lower Cheek Teeth MACN 5700 1.8 1.2 1.5 1.2 MACN 5701 1.2 1.5 1.1 MACN 5702 1.7 1.2 1.5 1.2 MACN 5703 1.8 1.1 1.5 1.2 1.3 1.0 0.4 0.6 5.6 MACN 5706 1.9 1.1 1.5 1.1 MACN 5708 1.3 1.0 MACN 5709 1.7 1.2 1.2 1.0 0.5 0.4 MACN 5710 1.8 1.2 1.5 1.1 1.1 1.0 MACN 5717 1.8 1.2 1.3 1.0 MACN 5720 1.6 1.3 1.3 1.0 MACN 8428 1.3 1.0 MACN 8429 1.6 1.2 1.3 1.0 0.6 0.5 MACN 8447 1.2 1.0 MACN 8448 1.8 1.2 1.6 1.2 1.3 0.9 ca5.7 MACN 8456 1.5 1.1 1.2 1.0 0.4 0.6 MACN 8464 1.9 1.2 1.6 1.2 1.2 1.0 0.4 0.4 5.7 MACN 8465 2.0 1.2 1.6 1.1 1.2 1.0 0.4 0.4 5.8 MACN 8466-7 2.0 1.2 1.8 1.2 1.3 1.0 0.5 0.5 5.8 Table 14. Statistics for some lower cheek tooth dimensions of Stilotherium dissimile. Dimens on N OR X s CV M, L 9 1.7-2.0 1.86 0.10 5.38 W 10 1.1-1.2 1.18 0.04 3.39 M2 L 15 1.5-1.8 1.57 0.09 5.73 W 15 1.1-1.3 1.17 0.06 5.13 M3 L 14 1.1-1.3 1.25 0.07 5.60 W 14 0.9-1.0 0.99 0.03 3.03 M, L 7 0.4-0.6 0.46 0.08 17.39 W 7 0.4-0.6 0.49 0.09 18.37 Ml^i W 5 5.6-5.8 5.72 0.08 1.40 an M4 and the breakage of the ascending portion of the ramus is similar, but not identical. The part of the figured jaw from P3 forward has no counterpart in the MACN collection.]; MACN 5717, a left mandibular ramus with P3 and M2_3 complete, roots of M, and M4 (syntype of H. didelphoides); MACN 5723, a fragment of a right mandibular ramus with P2_3 complete, base of M, and base of M2 trigonid (listed as "tipo" of S. dissimile in Ameghino's catalog); MACN 8426-29 and 8431 (all of a single, associated individual, labeled G. captiva) [8426, a frag- ment of a right maxillary with M2 3 (figured by Marshall, 1980, fig. 5); 8427, a fragment of a rostrum with left C-P2 and right P1 2 complete (figured by Marshall, 1980, fig. 4); 8428, a fragment of a right ramus with M3 complete and alveoli of M4; 8429, a left mandibular ramus with large incisor and M2_4 complete and alveoli of rest of dentition; and 8431, an isolated right M1 (figured by Marshall, 1980, fig. 5, erroneously as 8231)]; MACN 8447, a fragment of a right mandibular ramus with M3 complete, and alveoli of M4 (labeled G. minima); MACN 8448, a right mandibular ramus with large incisor, M,_3 complete, and alveoli of rest of i$&;. (Q O %; 2 U < £ 2 s in c || to ■d JU 3 79 68 MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 69 dentition (labeled G. minima); MACN 8456, a fragment of a left mandibular ramus with alveoli of M„ and M2_4 complete (labeled P. patagonicus and probably a cotype of that species, see Reig, 1955b, p. 64; he lists this specimen erroneously as 8450); MACN 8465, a fragment of a right mandibular ramus with P3-M4 complete [Part of the jaw figured by Ameghino (1897, fig. 75; 1903, figs. 33, 92; 1904b, fig. 32; 1906, fig. 196), the part from P3 posteriorly clearly was based on this specimen, as the breakage of the coronoid process is identical in the spec- imen and the illustration. The antemolar region and incisor may have been based on MACN 8464, but this is less certain.]; MACN 8466-67, a nearly complete left mandibular ramus with dentition; PU 15238, greater part of a left mandibular ramus with most of dentition, very worn, and various postcranial fragments (figured by Sinclair, 1906, pi. 63, figs. 8, 8a, 10-13). Horizon and locality. — All specimens are from the Santa Cruz Formation, Santa Cruz Province, Patagonia, southern Argentina and their localities and dates of collection are as follows: Sehuen MACN 5706, 5708, 5717 (collected by C. Ame- ghino, 1890-91); La Cueva MACN 8426-29, 8431, 8447, 8448, 8464, 8465, 8566-67 (collected by C. Ameghino, 1892-93); Monte Observacidn MACN 5698, 5700, 5701, 5702, 5703, 5709, 5710, 5720, 5723 (collected by C. Ameghino, 1890-91), MACN 8456 (collected by C. Ameghino, 1891-92); PU 15238 was collected by O. A. Peterson in 1896, five miles south of Coy Inlet. Age. — Santacrucian. Comments. — Among the syntypes of Halmarhiphus didelphoides, Garzonia annec- tens, Halmarhiphus nanus, and Phonocdromus patagonicus were included specimens referrable to Stilotherium dissimile and Microbiotherium patagonicum and/or M. te- huelchum. The taxonomic position of each of the syntypes of these species are discussed and clarified in the Systematic Section of the relevant microbiothere species. Phonocdromus Ameghino, 1894 Phonocdromus Ameghino, 1894, p. 355. Lectogenotype* of Phonocdromus. — P. gracilis8 Ameghino, 1894, p. 356. Diagnosis. — As for type and only known species. Known range. — As for type and only known species. Phonocdromus gracilis Ameghino, 1894. Figures 32, 33; Table 15. Phonocdromus gracilis Ameghino, 1894, p. 356; Reig, 1955b, p. 64, fig. 2; Marshall, 1980, p. 40, fig. 7. Garzonia gracilis Sinclair, 1906, p. 457. Halmarhiphus nanus Sinclair, 1906 (partim), p. 421, text fig. 7, pi. 63, figs. 9, 9a. Lectotype of P. gracilis. — MACN 8457, a fragment of a left mandibular ramus with P3-M3 complete (selected lectotype and figured by Reig, 1955b, p. 64, fig. 2; and Marshall, 1980, fig. 7). Hypodigm. — The type and MACN 5707, a fragment of a right mandibular ramus with M2_4 complete (a possible syntype of G. captiva); MACN 5722, a fragment of a left mandibular ramus with P3 complete, alveoli of Pi_2 and M,_4 (possible syntype of H. nanus); MACN 8458, a fragment of a right mandibular ramus with 8 Phonocdromus patagonicus was original genotype, but has been shown to be invalid by Reig (1955b, p. 64). 70 FIELDIANA: GEOLOGY Fig. 32. Phonocdromus gracilis Ameghino, 1894, p. 356 (Santacrucian). MACN 8457 (lectotype), a fragment of a left mandibular ramus with P3-M3 complete, and alveoli of M4: a, labial; b, occlusal; c, lingual views. Scale = 5 mm. Table 15. Measurements of lower cheek teeth of Phonocdromus gracilis. M, M: M, i M, Mr-4 Specimen L W L W L W L W L AMNH 9123 1.3 0.9 1.4 1.0 1.3 0.9 ca4.9 AMNH 9593 1.2 1.0 1.4 1.1 1.3 1.0 0.9 0.6 5.0 MACN 5707 1.5 1.0 1.2 0.8 0.6 0.5 MACN 8457 1.4 0.9 1.6 1.0 1.3 0.9 MACN 8458 1.5 1.0 1.3 0.9 ... 4.7 M2-3 complete and roots of M4 (labeled P. gracilis); MACN 8461, a fragment of a right mandibular ramus with alveoli of P3-M4 (labeled P. gracilis); MACN 8462, a fragment of a right mandibular ramus with M2_4 complete (labeled P. gracilis); AMNH 9123, a fragment of a left mandibular ramus with M,_3 complete and with alveoli of M4; AMNH 9593, a nearly complete right mandibular ramus with base of large incisor and with most of rest of dentition (figured by Sinclair, 1906, text fig. 7, pi. 63, figs. 9, 9a). Horizon and locality. — All specimens are from the Santa Cruz Formation, Santa MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 71 Fig. 33. Phonocdromus gracilis Ameghino, 1894, p. 356 (Santacrucian). AMNH 9593, a nearly complete right mandibular ramus with base of large incisor and with most of rest of dentition: a, labial; b, occlusal; c, lingual views. Scale = 5 mm. Cruz Province, Patagonia, southern Argentina and their localities and dates of collection are as follows: La Cueva MACN 8457, 8458, 8461, 8462 (collected in 1892-93); Sehuen MACN 5707 (collected in 1890-91); Monte Observacidn MACN 5722 (collected in 1890-91); all MACN specimens were collected by Carlos Ame- ghino. AMNH specimens were collected in 1899 by Barnum Brown along the north bank of the Rio Gallegos near Felton's Estancia. Age. — Santacrucian. Comments. — Among the syntypes of Halmarhiphus nanus were three specimens of which one is referrable to Stilotherium dissimile, one to Microbiotherium patago- nicum, and one to Phonocdromus gracilis (see p. 19). LITERATURE CITED Ameghino, F. 1887. Enumeraci6n sistem^tica de las especies de mamiferos f6siles colec- cionados por Carlos Ameghino en los terrenos eocenos de la Patagonia austral y de- positados en el Museo de La Plata. Bol. Mus. de La Plata, 1: 1-26. . 1889. Contribucion al conocimiento de los mamiferos f6siles de la Repiiblica Ar- gentina, obra escrita bajo los auspicios de la Academia Nacional de Ciencias de la Repiiblica Argentina para presentarla a la Exposicion Universal de Paris de 1889. Actas Acad. Cienc. C6rdoba, 6: 1-1027, atlas with 98 pis. . 1891. Nuevos restos de mamiferos f6siles descubiertos por Carlos Ameghino en el Eoceno inferior de la Patagonia austral. Especies nuevas, adiciones y correcciones. Rev. Arg. Hist. Nat., 1: 289-328. 72 FIELDIANA: GEOLOGY . 1893. Les premiers mammiferes. Relations entre les mammiferes diprotodontes eocenes de l'Amerique de Nord et ceux de la Republique Argentina. Rev. Gen. Sci., Paris, 4: 77-81. -. 1894. Enumeration synoptique des especies de mammiferes fossiles de formations eocenes de Patagonie. Bol. Acad. Cienc. Cordoba, 13: 295-452. -. 1897. Mammiferes cretaces de l'Argentine (Deuxieme contribution a la connaiss- ance de la faune mammalogique des couches a Pyrotherium). Bol. Inst. Geog. Argentino 18: 406-521. -. 1898. Sinopsis geologico-paleontologica. Segundo censo de la Republica Argentina. Fol., Buenos Aires, 1: 112-255. -. 1899. On the primitive type of the plexodont molars of mammals. Proc. Zool. Soc. Lond., 1899: 555-571. -. 1900. Mamiferos del Cretaceo inferior de Patagonia (Formacion de las areniscas abigarradas). Comun. Mus. Nac. Buenos Aires, 1: 197-206. 1902a. Cuadro sinoptico de las formaciones sedimentarias, terciarias y cretaceas de la Argentina en relacion con el desarrollo y descendencia de los mamiferos. An. Mus. Nac. Buenos Aires, (8), 1: 1-12. -. 1902b. Linea filogenetica de los proboscideos. An. Mus. Nac. Buenos Aires, 8, ser. 3, vol. 1: 19-42. -. 1902c. Sur le type primitif des molaires plexodontes des mammiferes. An. Mus. Nac. Buenos Aires, (3) 1: 419-39. -. 1902d. Notice preliminaires sur les mammiferes nouveaux des terrains cretaces de Patagonie. Bol. Acad. Cienc. Cordoba, 17: 5-70 (1902-04). 1902e. Premiere contribution a la connaissance de la faune mammalogique des couches a Colpodon. Bol. Acad. Cienc. Cordoba, 17: 71-141 (1902-04). 1903. Los diprotodontes del orden de los plagiaulacoideos y el origen de los roedores y de los polimastodontes. An. Mus. Nac. Buenos Aires, (3)1: 81-192. 1904a. Recherches de morphologie phylogenetique sur les molaires superieures des ongules. An. Mus. Nac. Buenos Aires, 10, ser. 3, vol. 3: 1-541. 1904b. Paleontologia argentina. Publicaciones de la Universidad de la Plata. Fa- cultad de ciencias fisico-matimaticas, 2: 1-79. 1904c. Nuevas especies de mamiferos cretaceos y terciarios de la Republica Ar- gentina. An. Soc. Cien. Argentina, 56: 193-208 (1903); 57: 162-75, 327-41 (1904); 58: 35-71, 182-92, 225-91 (1904). -. 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 1'ancien continent. An. Mus. Nac. Buenos Aires, (3), 8: 1-568. . 1907. El origen del hombre. La Reforma. La Plata, 1907: 256-65. 1909. Le Diprothomo platensis. Un precurseur de l'homme du Pliocene inferieur de Buenos Aires. An. Mus. Nac. Buenos Aires, (3), 12: 107-209. Bensley, B. A. 1903. On the evolution of the Australian marsupialia: with remarks on the relationships of the marsupials in general. Trans. Linn. Soc. Lond., (2), 9: 83-217. Biggers, J. D. 1966. Reproduction in male marsupials. In Comparative biology of repro- duction in mammals, I. W. Rowlands, ed., Symp. Zool. Soc. Lond. 15: 251-80. Biggers, J. D. and E. D. DeLamater. 1965. Marsupial spermatozoa pairing in the epi- didymis of American forms. Nature, 208: 402-4. Biggers, J. D., H. I. Fritz, W. C. D. Hare, and R. A. McFeely. 1965. Chromosomes of American marsupials. Science, 148: 1602-3. Cabrera, A. 1919. Genera Mammalium. Monotremata. Marsupialia. Museo Nac. de Cien- cias Nat., pp. 1-177. Madrid. Cabrera, A., and J. Yepes. 1960. Mamiferos Sud Americanos. Departamento de Publi- caciones Cientificas Argentinas, Buenos Aires, vols. 1 and 2 (new edition, first published in 1940). MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 73 Collins, L. R. 1973. Monotremes and marsupials: a reference for zoological institutions. Smithsonian Inst. Publ. 4888: 1-323. Corro, G. del. 1977. Un nuevo microbiotherio (Marsupialia) del Eoceno del Patagonia. Rev. Mus. Arg. Cien. Nat. "Bernardino Rivadavia," Paleont., 2(2): 31-3. Cunningham, R. O. 1871. Notes on the natural history of the Straits of Magellan and west coast of Patagonia, pp. 1-577. Edinburgh. Fischer, G. M. 1978. Los pequenos mamiferos de Chile. Gayana, Zoologia No. 40, Univ. de Concepcion, pp. 1-342. Gollan, J. S. 1946. La Comadrejita enana Dromiciops australis australis (F. Philippi). Holm- bergia 4: 191-97. Greer, J. K. 1965. Mammals of Malleco Province, Chile. Publ. Mus. Michigan State Univ. Biol. Ser. 3: 49-152. Hayman, D. L., and P. G. Martin. 1974. Animal cytogenetics. Chordata 4. Mammalia I: Monotremata and Marsupialia, pp. 1-110. Gebriider Borntraeger, Berlin, Stuttgart. Hill, J. P. and E. A. Fraser. 1925. Some observations of the female urogenital organs of the Didelphyidae. Proc. Zool. Soc. London 1925: 189-219. Hunsaker, D., II. 1977. Ecology of New World marsupials. In The Biology of Marsupials, D. Hunsaker, II, ed., pp. 95-196. Academic Press, New York. Kirsch, J. A. W. 1977. The comparative serology of Marsupialia, and a classification of marsupials. Austral. J. Zool., suppl. ser., no. 52: 1-152. Kirsch, J. A. W. and J. H. Calaby. 1977. The species of living marsupials. In The Biology of Marsupials, B. Stonehouse and D. Gilmore, eds., pp. 9-26. Univ. Park Press, London. Krieg, H. 1924. Biologische Reisestudien in Sudamerika. III. Chilenische Beutelratten. Zeitsch. Morphol. Okol. der Tiere, Berlin, 3: 169-76. Mann, G. F. 1955. Monito del monte Dromiciops australis Philippi. Invest. Zool. Chilenas, 2: 159-66. . 1958. Reproducion de Dromiciops australis (Marsupialia: Didelphydae). Invest. Zool. Chilienas, 4: 209-13. Marshall, L. G. 1978. Dromiciops australis. Mammalian Species, American Society of Mammalogists, 99: 1-5. . 1979. Evolution of metatherian and eutherian characters: a review based on cladistic methodology. Zool. J. Linn. Soc. Lond., 66: 369-410. . 1980. Systematics of the South American marsupial family Caenolestidae. Field- iana: Geology, n.s. no. 5: 1-145. Marshall, L. G., R. Hoffstetter, and R. Pascual. In Press. Geochronology of the mammal-bearing Tertiary of South America. Fieldiana: Geology. Osgood, W. H. 1943. The mammals of Chile. Field Mus. Nat. Hist., Zool. Ser. 30: 1-268. Pascual, R. and H. E. Herrera 1975. Stilotherium Ameghino, 1887, el mas primitivo Caenolestidae conocido. Consideraciones sobre la transicion Didelphidae-Caenolestinae (Marsupialia). Act. 1st. Congr. Argentino, Paleont. y Bioestr., Tucum^n, 1974, 2: 417-30. Pascual, R. and O. E. Odreman Rivas. 1971. Evolucion de las comunidades de los vertebrados del Terciario Argentino. Los aspectos PaleozoogeogrSficos y Paleoclimaticos relacionados. Ameghiniana, 8: 372-412. Patterson, B. 1965. The auditorv region of the borhyaenid marsupial Cladosictis. Breviora, 217: 1-9. Philippi, F. 1893a. Un nuevo marsupial chileno. Anal. Univ. Chile, 86: 31-4. . 1893b. Ein neues Beutelthier Chiles. Verhandl. Deutsch. Wiss. Verein., Santiago, Cantiago de Chile, 5: 318-19. Reig, O. A. 1955a. Noticia preliminar sobre la presencia de Microbiotherinos vivientes en la fauna Sudamericana. Invest. Zool. Chilienas, 2: 121-30. . 1955b. Un nuevo genero y especie de cenolestinos del Plioceno de la Provincia de Buenos Aires (Repiiblica Argentina). Rev. Asoc. Geol. Argent., 10: 60-71. In Press. Teoria del origen y desarrollo de la fauna de mamiferos de America del Sur. Public. Mus. Munic. de Cienc. Natur. "Lorenzo Scaglia," Mar del Plata. 74 FIELDIANA: GEOLOGY Reig, O. A. and G. G. Simpson. 1972. Sparassocynus (Marsupialia, Didelphidae), a peculiar mammal from the late Cenozoic of Argentina. J. Zool. 167: 511-39. Reig, O. A., J. A. W. Kirsch and L. G. Marshall. In Prep. Systematic relationships of extant and Neocenozoic didelphoid marsupials. Ringuelet, A. B. de. 1953. Revision de los didelfidos fosiles argentinos. Revta. Mus. Univ. Eva Peron (n.s.), La Plata, Secc. Paleont., 3: 265-308. Rusconi, C. 1933. New Pliocene remains of diprotodont marsupials from Argentina. Jour. Mammal., 14: 244-50. Schaeffer, B. 1947. An Eocene Serranid from Patagonia. Amer. Mus. Novit., 1331: 1-9. Schlosser, M. 1923. Grundzuge der Paleontologie (Palaeozoologie) von Karl A. von Zittel, II Abteilung — Vertebrata. Neuarbeitet von F. Broili und M. Schlosser, R. Oldenbourg, Munich and Berlin, 706 pp. . 1925. Class V. Mammalia, Vol. III. Textbook of Paleontology by K. A. von Zittel, revised by Max Schlosser, MacMillan & Co., London, 316 pp. Scott, W. B. 1937. A history of Land Mammals in the Western Hemisphere. Revised edition. The MacMillan Co., New York, 786 pp. Segall, W. 1969a. The auditory ossicles (malleus, incus) and their relationships to the tympanic in marsupials. Acta Anat. 73: 176-91. . 1969b. The middle ear region of Dromiciops. Acta Anat. 73: 489-501. . 1970. Morphological parallelisms of the bulla and auditory ossicles in some in- sectivores and marsupials. Fieldiana: Zool. 51: 169-205. Sharman, G. B. 1973. The chromosomes of non-Eutherian mammals. In Cytotaxonomy and vertebrate evolution, A. B. Chiarelli and E. Capanno, eds., pp. 485-530. Academic Press, London. Simpson, G. G. 1929. American Mesozoic Mammalia. Mem. Peabody Mus., 3 (1): 1-171. . 1930. Post-Mesozoic Marsupialia. In Fossilium Catalogus. 1: Animalia. Berlin, W. Junk, Pars. 47: 1-87. 1932a. The supposed occurrences of Mesozoic mammals in South America. Amer. Mus. Novit., 530: 1-9. . 1932b. Some new or little-known mammals from the "Colpodon" beds of Patagonia. Amer. Mus. Novit., 575: 1-12. . 1935. Note on the classification of Recent and fossil opossums. Jour. Mamm., 16(2): 134-37. 1938. A new marsupial from the Eocene of Patagonia. Amer. Mus. Novit., 989: 1-5. . 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 I. Bull. Amer. Mus. Nat. Hist., 91: 1-232. 1964. Los mamiferos Casamayorenses de la Coleccion Tournouer. Rev. Mus. Arg. Cien. Nat. Paleont., 1: 1-21. 1967. The Beginning of the Age of Mammals in South America. Part II. Bull. Amer. Mus. Nat. Hist., 137: 1-259. Sinclair, W. J. 1906. Marsupials of the Santa Cruz beds. Repts. Princeton Univ. Expeds. Patagonia, 4: 330-460. Spotorno, O. A. and D. R. Fernandez. 1971. The chromosomes of the "Monito del Monte", Dromiciops australis. Mammal. Chromosome Newsl., 12: 40. Stoll, N. R., et al. 1961, 1964. International Code of Zoological Nomenclature Adopted by the XV International Congress of Zoology. London, International Trust for Zoological Nomenclature, pp. 1-176. Tate, G. H. H. 1933. A systematic revision of the marsupial genus Marmosa. Bull. Amer. Mus. Nat. Hist. 66: 1-250. Thomas, O. 1894. On Micoureus griseus, Desm., with the description of a new genus and species of Didelphyidae. Ann. Mag. Nat. Hist., Ser. 6, 14: 184-88. MARSHALL: SYSTEMATICS OF MICROBIOTHERIIDAE 75 Tournouer, A. 1903. Note sur la geologie et la paleontologie de la Patagonie. Bull. Soc. Geol. France, 4 ser., 3: 463-73. Walker, E. P., et al. 1963. Mammals of the World. Johns Hopkins Press, Baltimore, 2nd ed. 2 vols., 1500 pp. Winge, H. 1923. Pattedyr-Slaegter. I. Monotremata, Marsupialia, Insectivora, Chiroptera, Edentata. Copenhagen, viii + 360 pp., Ill pis. [English Translation: "The Interrela- tionships of the Mammalian Genera." Copenhagen, 3 vols. (1941-1942). Transl. by E. Deichmann and G. M. Allen]. Wolffsohn, J. A. 1913. Contribuciones a la mammalogia chilena: Apuntes sobre los mar- supiales. Rev. Chil. Hist. Nat., 17: 74-88. Field Museum of Natural History Roosevelt Road at Lake Shore Drive Chicago, Illinois 60605-2496 Telephone: (312) 922-9410