UBUVERS" - ■JUWDIS i ^ vr URBANACHAMPAIGN ClOLOGY % FIELDIANA Geology Published by Field Museum of Natural History Volume 33, No. 18 March 31, 1976 This volume is dedicated to Dr. Rainer Zangerl The Brain of Mesonyx, A Middle Eocene Mesonychid Condylarth LEONARD RADINSKY Department Of Anatomy University Of Chicago INTRODUCTION fHS y£**SC QE IHf The Mesonychidae is a family of medium-sized to gigantic; am^^p^s and carnivores that existed during the Paleocene and Eocene epochs in North America, Europe, and Asia. Mesonychids were foiTnerljjpteffii^llVvlQ^lbttWOJS donts as archaic members of the Order Carnivora (Simp^MRr^J?*Sfia most earlier workers), but recently have been reassigned to the Order Condylar- thra, to better reflect phylogenetic relationships (Van Valen, 1966; Romer, 1966). Condylarths were a heterogeneous group of early Tertiary, predomi- nantly small to medium-sized omnivores and herbivores, from which the various ungulate and subungulate orders were derived. For an introduction to the literature on mesonychids, see Szalay and Gould (1966) and Szalay (1969). Endocranial casts of representatives of most of the families of condy- larths have been described. These are of the arctocyonids Arctocyonides and Arctocyon (Russell and Sigogneau, 1965); periptychid Periptychus (Tilney, 1931; Edinger, 1956); hyopsodontid Hyopsodus (Gazin, 1968); phenacodon- tid Phenacodus (Tilney, 1931; Simpson, 1933); meniscotheriids Pleuraspido- therium (Russell and Sigogneau, 1965) and Meniscotherium (Gazin, 1965); and the tillodontid Tillodon (Gazin, 1953). Scott (1888) described a partly exposed natural endocast of Mesonyx but his few observations, unsupported by figures or measurements, provide no useful information. The endocast of Mesonyx described below is important because it provides the first good Library of Congress Card Number: 75-27501 Publication 1226 323 [ 5 cm &$p&^?VV\ jVs Fig. 1. Mesonyx obtusidens, Yale Peabody Mus. 13141. Above, dorsal and lateral views of endocast, approximately natural size. Below, lateral view of endocast in position in skull, approximately x 2/5. Dashed lines indicate estimated boundaries of missing portions; dotted iines indicate border of unexposed portions of endocast. 324 RADINSKY: BRAIN OF MESONYX 325 record of a mesonychid brain, and because it is one of the largest and one of the latest condylarth endocasts known. DESCRIPTION A natural endocast was exposed by removing most of the right side of the braincase of Yale Peabody Mus. 13141, a well-preserved, uncrushed skull of Mesonyx obtusidens, from the Middle Eocene (Bridger B, about 50 million years old), of Wyoming. The exposed portion of the endocast (fig. 1) includes most of the cerebellum, cerebrum, and olfactory bulbs. The rhinal fissure is located about two-thirds of the way down the cerebrum, as seen in lateral view. It is well-marked caudally, but rostrally is faint. A cast of a vascular sinus overlies its middle portion, a condition commonly seen in other mammals. Three longitudinally oriented sulci divide the neocortex above the rhinal fissure. The most lateral of these sulci is unusually long and straight compared to what is seen in other mammals, extending for most of the length of the hemisphere. The middle sulcus is shorter, with a faint suggestion of a bifurcation at its rostral end. The medial sulci is almost as long as the bottom one, and it appears to curve slightly laterally at its rostral end. Enough is preserved of the olfactory chamber to indicate that the olfac- tory bulbs were slightly pedunculate and relatively small. The pyriform lobe also appears to have been relatively small compared to the rest of the brain (see fig. 2 for comparison with early mammals). The midbrain was not completely overlapped, for there is a gap of about 5 mm. between the caudal end of the cerebrum and the cerebellar vermis. No details of midbrain morphology are preserved in that space. The vermis is clearly demarcated from the lateral hemispheres, but otherwise little sur- face detail of cerebellar morphology is evident. The vermis is relatively high and short, with a transverse groove located relatively rostrally. That groove may represent the fissura prima, for on most mammal endocasts the / prima is the most prominent, and often the only cerebellar fissure reproduced. There is a faint indication of a longitudinally oriented goove on the side of the cerebellar hemisphere; it may represent the boundary between the an- siform lobule and the paraflocculus. The cerebellar hemispheres extend out about as far laterally as the cerebral hemispheres. From water displacement of a cast of the Mesonyx endocast, with olfac- tory bulbs and the covered portion of the hind brain restored, I estimate the endocranial volume to have been about 80 cc. 326 FIELDIANA: GEOLOGY, VOLUME 33 I have exposed the cerebrum of the Mesonyx endocast described by Scott (1888), Princeton Univ. 10308. It is somewhat crushed and incomplete, but appears similar in observable details to the Yale specimen. I see no basis for Scott's description of the cerebral hemispheres as very small and the cerebel- lum as relatively large. MORPHOLOGICAL COMPARISONS Factors to consider in comparisons with brains of other mammals are body size, phylogenetic relationship, temporal relationship (geological age), and ecological niche. Most closely related phylogentically to Mesonyx are representatives of the other condylarth families. Of these, Arctocyonides, Hyopsodus, Pleuraspidotherium, and Meniscotheriwn were considerably smaller than Mesonyx. Therefore, the fact that their cerebral hemispheres lacked convolutions or at most had a single neocortical sulcus (see references cited above) does not necessarily indicate a less advanced stage of cortical evolution than in Mesonyx, since degree of cortical folding in some groups of mammals appears to be at least in part correlated with absolute brain size, which, in turn, is correlated with body size. The influence of size on degree of gyrencephaly can be seen in series of brains of living prosimian primates (Radinsky, 1974) and ceboid primates (Hershkovitz, 1970). However, such influence does not appear as evident in cercopithecid primates (Connolly, 1950; Radinsky, pers. observation) or in canid carnivorans (Radinsky, 1973). Of the remaining condylarths for which endocasts are known, Peripty- chus, Arctocyon, Phenacodus, and Tillodon were closer in body size to Meso- nyx, although somewhat smaller. The endocast of Tillodon is crushed and does not preserve enough surface detail for significant comparison with the other genera. The brain of Periptychus, from the Middle Paleocene (about 65 million years old), is known from the dorsal half of an endocast of the fore brain. It has a very high rhinal fissure, and only two small caps of neocortex on top of the cerebrum. The brain of Arctocyon (fig. 2 A), from Opposite: Fig. 2. Drawings of endocasts of condylarths (A, B and C), an early ungulate (D), and early carnivorans (E and F), in dorsal, lateral, and rostral views. See text for discussion. Dashed lines indicate estimated boundaries of missing portions. A, Arctocyon primaevus, redrawn from Russell and Sigogneau, 1965; B, Phenacodus primaevus, redrawn from Simpson, 1933; C, Mesonyx obtusidens, Yale Peabody Mus. 13141; D, Hyrachyus modestus, Amer. Mus. Nat. Hist. 11713, with cerebellum restored from other specimens; E, Hyaenodon horridus, Amer. Mus. Nat. Hist. 94760; F, Humbertia angustidens, Mus. Nat. Hist. Nat., Paris, from a cast of the original specimen. Abbreviations: f, fissura prima; r, rhinal fissure. All drawings to same scale, ?bout r. \/2. A. Arc t oc yon B. P h enacodus C. M e s o n y x D. H y r a c h y u s F. H u m bertia E H y a e n o d on 327 328 FIELDIANA: GEOLOGY, VOLUME 33 the Late Paleocene (about 60 million years old), was advanced over that of Periptychus in having relatively more neocortex, evidenced by a slightly lower rhinal fissure and the presence of one or possibly two neocortical sulci (surface details are poorly defined on the two known Arctocyon endocasts). The midbrain was widely exposed in Arctocyon. The brain of Phenacodus (fig. 2B), from the Early Eocene (about 55 million years old), was further ad- vanced in having a lower rhinal fissure, with the neocortex covering more of the midbrain and olfactory peduncles than in Arctocyon. One or possibly two neocortical sulci were present in Phenacodus. (As in the cast of Arcto- cyon, surface details are poorly preserved on the one described endocast of Phenacodus.) The brain of Mesonyx was advanced over those of the above mentioned condylarths in having a relatively more expanded neocortex. This is indicated by the relatively more ventrolateral position of the rhinal fissure, the presence of three well-defined neocortical sulci, and the expansion of the cere - brum above the height of the cerebellum. In addition, if the transverse groove on the cerebellar vermis represents the Jlssura prima, it is in a more rostral position than in the other condylarth endocasts, indicating expansion of the neocerebellar portion of the vermis, a progressive trend presumably correlated with the expansion of the neocortex of the cerebrum. The olfactory bulbs are relatively smaller in Mesonyx than in the other condylarths in which their size is known. Finally, even allowing for the more ventrally located rhinal fissure, the pyriform lobe appears to have been relatively smaller in Mesonyx than in the other condylarths. There are no other condylarth endocasts of the same geological age or younger than the Mesonyx endocast with which it may be compared. The next most closely related group for which Middle Eocene endocasts are known is the ungulate order Perissodactyla, which evolved from phenaco- dontid condylarths in the Middle or Late Paleocene. Hyrachyus, a helaletid tapiroid, was comparable in size to and contemporaneous with Mesonyx and therefore suitable for comparison of external brain morphology. The brain of Hyrachyus (fig. 2D) appears to have been similar in overall proportions and degree of neocortical expansion to those of other Middle Eocene per- issodactyls. Although the rhinal fissure was not as ventrally located in Hyra- chyus as in Mesonyx, its brain also had three well developed neocortical sulci and a fourth short one. However, the rostral end of the most lateral sulcus curved medially and delimited a portion of frontal cortex in Hyrachyus that is not so bounded in Mesonyx. The brain of Hyrachyus further differed from that of Mesonyx in having less reduced olfactory bulbs and pyriform lobe, a more caudally located fissura prima, and in being narrower across the cerebellum than across the cerebrum. Thus in degree of neocortical expan- RADINSKY: BRAIN OF MESONYX 329 sion, brains of Mesonyx and Hyrachyus appear to have been similar, although there are differences in details (e.g., lower rhinal fissure in Mesonyx and more differentiated frontal pole in Hyrachyus). The more rostrally located fissura prima suggests that the cerebellum of Mesonyx was more advanced than that of Hyrachyus. The relatively smaller olfactory bulbs and pyriform lobe of Mesonyx are also specialized features. While mesonychids are phylogenetically closer to condylarths and un- gulates than to carnivorans, in general habitus they appear to have been more similar to the latter, particularly during Eocene times (see Szalay and Gould, 1966). Therefore, it is of interest to compare the Mesonyx endocast with those of early carnivorans. Archaic carnivorans, called creodonts, unrelated to the ancestry of modern carnivorans, were abundant during the Eocene. However, the earliest known creodont endocasts from animals close in size to Mesonyx are from the Oligocene, about 15 million years later in time. The brain of Hyaenodon horridus (fig. 2E), a hyaenodontid creodont, was more advanced than the other known Eocene and Oligocene creodont brains. It had two major neocortical sulci, and two shorter, variably developed ones. The lower major sulcus curved medially at its rostral end, as in the Eocene perissodactyls and unlike the straight lower sulcus in Mesonyx. However, despite the expansion of the neocortex indicated by the presence of so many sulci, the rhinal fissure in Hyaenodon was not as ventrally displaced as in Mesonyx. Also, the olfactory bulbs and pyriform lobe are relatively larger and the cerebellar fissura prima apparently less rostrally displaced in Hyaeno- don than in Mesonyx. The modern families of carnivorans, or neocarnivorans, appear to have arisen from a late Eocene adaptive radiation of miacid carnivorans. Judging from the various known Oligocene neocarnivoran endocasts (see Piveteau, 1951; Radinsky, 1971, 1973), the basal neocarnivoran brain was probably similar to that of Humbertia angustidens (described under the name Viver- ravusby Piveteau, 1962), a late Eocene miacid. The brain of Humbertia (fig. 2F) resembled that of Mesonyx in the position of the rhinal fissure, but differed in having only two neocortical sulci (the coronolateral and suprasyl- vian sulci), with a wide unfolded area of cortex between the lower sulcus and the rhinal fissure. Unlike the conditions in Mesonyx, the sulci in Humbertia are gently arched; in later carnivorans, the arching becomes even more pronounced. Olfactory bulbs were relatively larger in Humbertia and the fissura prima less rostrally displaced than in Mesonyx. Because of the differ- ence in overall brain size, it is difficult to estimate the relative size of the pyriform lobe in Humbertia compared to Mesonyx. The midbrain is com- pletely covered by the cerebrum in Humbertia. 330 FIELDIANA: GEOLOGY, VOLUME 33 Because Humbertia was considerably smaller than Mesonyx, it would be of interest to determine to what degree allometry was responsible for the observed differences in brain morphology, particularly in the number of neocortical sulci. Oligocene neocarnivorans that were closer in size to Meso- nyx, such as the amphicyonids Amphicyon and Daphoenus, and the canid Mesocyon, had brains that were similar in most features to that of Humbertia, but had in addition a third sulcal arch, the ectosylvian sulcus, beneath a more convex suprasylvian sulcus (Beaumont, 1964; Radinsky, 1971, 1973). Small early neocarnivorans generally lack an ectosylvian sulcus, which suggests that its absence in the Late Eocene Humbertia might be due to allometry. In addition to the ectosylvian sulcus, some but not all large early neocarnivo- rans have one or two short secondary sulci, the ectolateral and entolateral sulci, adjacent to the caudal end of the cornolateral sulcus. RELATIVE BRAIN SIZE Most of the statements in the literature on relative brain size of fossil mammals are unsupported assertions that brains were relatively small in any given extinct species. For example, Scott (1888, p. 155) wrote that Mesonyx had an exceedingly small brain capacity, but did not specify his point of comparison, and gave no measurements of endocranial capacity. However, during the past 10 years, Jerison (1973 and references cited therein) has provided a large body of quantitative data on relative brain size in fossil mammals, based on endocranial volumes (used interchangeably with brain weights) and body weights (estimated from various skeletal measurements). For purposes of comparison, Jerison uses an Encephalization Quotient, or EQ, which is defined as the endocranial volume (or brain weight) of a given species divided by the endocranial volume one would expect to find in an "average" living mammal of that species' body weight. The relationship between brain weight and body weight in Jerison's "average" living mammal is described by the equation, E = 0.12 P0-67, (E = brain weight; P = body weight), based on a large sample of living mammals. Bauchot and Stephan (1966) compare relative brain sizes of recent mammals in a similar manner, except their Encephalization Index is based on a comparison with the brain size one would expect in a basal insectivore of a given body weight. For brain weight-body weight comparisons, I would have preferred to use Bauchot and Stephan's basal insectivore line as a standard, since the equation describing relative brain size in basal insectivores is unlikely to change with the addition of more data, while the equation for the "average" living mammal is more likely to vary depending on what species are included in the sample. However, since Jerison has calculated EQs for a large number of fossil ■e C O U X> OS < CM 1—1 00 or 1— ( d w C - o ^~- •XSO* 5 w N ^ ~ -(J t> CD cn es c 1/5 Tt CO J3 . iJ. co 'C ^ W ° PQ *H -(J lie GO ^ w PQ B * cu o CO s S 8 I C O w > <« X O (D +J C8 0) e cu i cct -<-> e be be '53 a £ _> >> ~^ T3 O "cu .2 be cS C >H CO a> J2. > -u rt t-i a> c CO o s ft 3 X c cu be ed S 'co c C cu '3 e Sh cS ,£ u X i.2 T3 aJ > <£ C O CO •s Cfl ft CO >-. e cu a> o "8 -fl 0 1 § co «C c o C o T3 cu > , CO •Gex cs^ -u ° CO C cu '•<3 •3* O D c co be C 0 cu o c CS E a x-° a c -4J m co X, >- CO cu ft a> cu JZ — X 0 c B w --E 331 332 FIELDIANA: GEOLOGY, VOLUME 33 mammals, to facilitate comparisons with his data I have used EQ to provide a quantitative measure of relative brain size in Mesonyx. Data for estimations of relative brain size in Mesonyx and other condy- larths for which such data are available are presented in Table 1 . Endocranial volumes were measured by water displacement of endocasts, (or copies of endocasts), with distorted or missing parts restored. My estimate of endo- cranial volume Arctocyonides differs from that of Jerison (1973, Table 11.1). Body weight was estimated from body length, using the equation P=.025 L3, which was calculated by Jerison (1973, p. 53) from a large sample of living ungulates and carnivores. For Arctocyon and Arctocyonides, Jerison used the equation P = .050 L3 to calculate body weight, which resulted in lower EQs than I calculated, but from the known skeletal remains of those genera I see no reason not to use the same equation as for the other condylarths. Body length (skull and trunk length) could be measured directly only in Phenaco- dus; for the other genera I estimated it from the proportions of skull length to body length in other specimens or in related genera. The Encephalization Quotient for Mesonyx obtusidens is 0.46 or 0.57, depending on the estimate of body length used. EQs for the other condylarths ranged from 0.20 to 0.32, all under the Mesonyx minimum estimate. For Middle Eocene perissodactyls, Jerison (ibid.) estimated three EQs, ranging from 0.37 - 0.49; for the Hyrachyus specimen shown in Figure 2D, I estimated an EQ of 0.36 or 0.52, depending on the body length estimate used. For four Eocene and Oligocene creodont carnivores, Jerison (ibid.) calculated EQs ranging from 0.33 - 0.55; for my Hyaenodon horridus specimen, I estimate an EQ of 0.61. For a sample of early neocarnivorans, Jerision calculated EQs ranging from 0.32-0.79. Thus, compared to contemporaneous perissodactyls and to creodonts and early neocarnivorans, Mesonyx did not have a relatively small brain. For comparison of relative brain size of Mesonyx with modern species, I calculated mean EQs and the observed range of EQs for representative samples of living insectivorans (data from Bauchot and Stephan, 1966), of artiodactyls (the dominant surviving ungulates), and of carnivorans (see table 2). Relative brain size of Mesonyx was in the upper part of the observed range of relative brain size of living insectivorans, in the lower part of the observed range of living artiodactyls, and around the lower end of the ob- served range of living carnivorans. One of the problems with analyzing relative brain size by the above method is the uncertainty involved in estimating body weight of extinct species from skeletal measures. Even where complete skeletons are available, and body length can be measured directly, it is evident from the graph presented by Jerison (1973, p. 53, fig. 2.9) that there is a high degree of RADINSKY: BRAIN OF MESONYX 333 variability in the body length-body weight relationship. Unfortunately, Jeri- son does not provide calculations of the variance, but it appears from his graph that for a mammal of 100 cm. body length, the observed range of body weight is from 10 kg. to about 35 kg. This problem might be minimized if one could calculate EQs for several related, approximately contemporaneous species, for errors of body weight estimates would probably be random and therefore with a large enough sample would cancel each other out. However, Mesonyx is the only mesonychid for which brain size can be estimated, and there are not even any other condylarths of Middle Eocene age available with which it can be compared. Therefore, it is desirable to have another method of estimating relative brain size to provide a check on the Encephalization Quotient calculated for Mesonyx. Plots of brain weight vs. foramen magnum area for six groups of living mammals (insectivorans, rodents, prosimian primates, artiodactyls, carnivo- rans, and monkeys) show approximately the same relative relationships as do brain weight-body weight plots of those groups (Radinsky, 1967). For a sample of 164 recent mammal species of those six groups, the coefficient of correlation (r) between foramen magnum area and body weight is 0.98. Removing the influence of brain weight, the partial correlation of foramen magnum area and body weight is 0.65. Therefore, it seems reasonable to Table 2. Relative brain size in some living mammals. Encephalization Quotients1 EQA2 Order Mean Observed Range Mean Observed Range 0.41 0.20-0.68 1.08 0.69-1.52 Insectivora3 0.47 0.24-0.83 (N = 24) Artiodactyla4 0.81 0.39-1.29 (N = 36) Carnivora' 0.89 0.52-1.80 (N = 48) 1.37 0.88-2.62 1 EQ is the brain size of a given species divided by the brain size expected for an "average" living mammal of that species' body weight. See text for further information. 2 Encephalization quotient based on comparison with foramen magnum area rather than body weight. See text for further information. 3 Brain weight and body weight data from Bauchot and Stephan, 1966. 4 Body weight data from Kruska, 1973, and Walker, 1964. 5 Body weight data from Walker, 1964. 334 FIELDIANA: GEOLOGY, VOLUME 33 examine the relationship between brain size and foramen magnum area in Mesonyx as a check on the relative brain size as estimated from body weight. To facilitate comparisons, I calculated the equivalent of the EQ for foramen magnum data. The EQA of a given species is the observed brain size of that species divided by the brain size one would expect in an "average" living mammal of that species' foramen magnum area. In my sample of 164 species of insectivorans, rodents, prosimians, articdactyls, carnivorans, and mon- keys, the average brain weight-foramen magnum area relationship is ex- pressed by the equation E = 22.4 A1-48, or log E = 1.35 + 1.48 log A (A = foramen magnum area, cm2). The results of this approach are presented in Tables 1 and 2. Relative brain size in Mesonyx based on the foramen magnum area comparison is higher than that of Arctocyonides, Meniscotherium, and Phenacodus, the other condylarths for which the relevant data are available, and comparable to that of early perissodactyls and carnivorans (Radinsky, unpublished data). This confirms the analysis based on body weight com- parisons. However, compared to the recent species, EQAs of Mesonyx and the other condylarths are higher relative to their EQs. Thus on the basis of foramen magnum area, relative brain size in Mesonyx is above the observed range of insectivorans, just above the mean for artiodactyls, and well within the lower part of the observed range for carnivorans. Two possibilities to account for this difference are that we have overestimated body weights for the extinct genera (and thus have EQs that are too low), or that the relation- ship between foramen magnum area and body weight is different in the living species compared to the fossil ones. CONCLUSIONS The brain of Mesonyx was relatively larger and more advanced in terms of expansion of neocortex (and probably also neocerebellum) than the other known condylarth brains with which it may be compared. The latter, however, are from earlier time periods than Mesonyx. The brain of Mesonyx was roughly comparable in relative size and in degree of neocortical expan- sion compared to brains of contemporaneous perissodactyl ungulates and slightly younger (geologically) carnivorans. The cerebellar fissura prima ap- pears to be rostrally displaced in Mesonyx compared to early ungulates and carnivorans, suggesting a relatively more expanded neocerebellum. The brain of Mesonyx was more specialized than that of other condy- larths and of early ungulates and carnivorans, in having relatively small olfactory bulbs and apparently a relatively smaller pyriform lobe. The rela- tive size of the pyriform lobe is difficult to estimate, and its apparent reduc- RADINSKY: BRAIN OF MESONYX 335 tion in Mesonyx may in part be an illusion resulting from the relatively great expansion of the neocortex. If the pyriform lobe was indeed relatively small in Mesonyx, that may be correlated with the reduction of the olfactory bulbs, since the pyriform lobe cortex is usually considered to be mainly involved in olfactory function. The rhinal fissure was more ventrally displaced in Mesonyx than in the early ungulates and carnivorans that had a similar number of neocortical sulci, suggesting either a greater degree of neocortical expansion in Mesonyx or that its sulci were shallower. The sulcal pattern of Mesonyx is so different from that of ungulates and carnivorans (or of any other mammal), that I hesitate to attempt to identify sulci and interpret functional areas of the cortex. The only functional inter- pretation that is apparent from the known brain morphology of Mesonyx is reduction in importance of olfaction, indicated by the apparently reduced olfactory bulbs. I see no features of the brain of Mesonyx that suggest phylo- genetic affinity to any other group of mammals. ACKNOWLEDGEMENTS For permission to prepare and describe the endocast from the Yale Peabody Museum Mesonyx skull, I am grateful to Prof. E. Simons. For access to other specimens utilized in this study, I thank D. Baird, Princeton Univer- sity; C. Dechaseaux, Museum National d'Histoire Naturelle, Paris; C. L. Gazin, U. S. National Museum; and M. McKenna and R. Tedford, the American Museum of Natural History. This work was supported in part by National Science Foundation Grant GB 31242. REFERENCES Bauchot, R. and H. Stephan 1966. Donnees nouvelles sur l'encephalisation des insectivores et des prosimiens. Mammalia, 30, pp. 160-196. Beaumont, G. 1964. Un crane d'Amphicyon ambiguus (Filhol) (Carnivora) des Phosphorites du Quercy. Arch. Sci. Geneve, 17, pp. 331-339. Connolly, C. J. 1950. External morphology of the primate brain. C. C. Thomas, Springfield, Illi- nois, 378 pp. Edinger, T. 1956. Objets et resultats de la paleoneurologie. Ann. Paleontol., 42, pp. 97-1 16. 336 FIELDIANA: GEOLOGY. VOLUME 33 Gazin. C. L. 1953. The Tillodontia: An early Tertiary order of mammals. Smithson. Misc. Coll., 121(10), pp. 1-110. 1965. A study of the early Tertiary condylarthran mammal Meniscotherium. Smithson. Misc. Coll., 149(2), pp. 1-98. 1968. A study of the Eocene condylarthran mammal Hyopsodus. Smithson. Misc. Coll., 153(4), pp. 1-90. Hershkovitz, P. 1970. Cerebral fissural patterns in platyrrhine monkeys. Folia Primat., 13, pp. 213-240. JERISON, H. 1973. Evolution of the brain and intelligence. Academic Press, New York, 482 pp. Kruska. D. 1973. Cerebralisation, Hirnevolution und domestikationsbedingte Hirngrdssen- anderungen innerhalb der Ordnung Perissodactyla Owen, 1848 und ein Ver- gleich mit der Ordnung Artiodactyla Owen, 1848. A. zool. Systematik Evolu- tions forschung, 11, pp. 81-103. PlVETEAU. J. 1951. Recherches sur revolution de l'encephale chez les carnivores fossiles. Ann. Paleontol.,37,pp. 133-151. 1962. L'encephale de Viuerravus angustidens, miacide des Phosphorites du Quercy. Ann. Paleontol., 48, pp. 163-175. Radinsky, L. 1967. Relative brain size: a new measure. Science, 155, pp. 836-837. 1971. An example of parallelism in carnivore brain evolution. Evolution, 25, pp. 518-522. 1973. Evolution of the canid brain. Brain, Behavior, Evol., 7,(3), pp. 169-202. 1974. Prosimian brain morphology: Functional and phylogenetic implications. In Doyle, G. A., R. D. Martin, and A. Walker, eds. Proceedings of the research seminar on prosimian biology, London, 1972. Duckworth, London. Romer, A. S. 1966. Vertebrate Paleontology. Univ. Chicago Press, Chicago, 468 pp. Russell, D. E. and D. Sigogneau 1965. Etude de moulages endocraniens de mammiferes Paleocenes. Mem. Mus. Nat. d'Hist. Nat., ser. C, 16, pp. 1-35. Scott, W. B. 1888. On some new and little known creodonts. J. Acad. Nat. Sci. Philadelphia, 2nd ser., 9(2), pp. 155-185. Simpson, G. G. 1933. Braincasts of Phenacodus, Notostylops, and Rhyphodon. Amer. Mus. Nov., no. 622, pp. 1-19. 1945. The principles of classification and a classification of mammals. Bull. Amer. Mus. Nat. Hist., 85, pp. 1-350. Szalay, F. S. 1969. The Hapalodectinae and a phylogeny of the Mesonychidae (Mammalia, Condylarthra). Amer. Mus. Nov., no. 2361, pp. 1-26. RADINSKY: BRAIN OF MESONYX 337 Szalay, F. S. and S. J. Gould 1966. Asiatic Mesonychidae (Mammalia, Condylarthra). Bull. Amer. Mus. Nat. Hist., 132(2), pp. 127-174. TlLNEY, F. 1931. Fossil brains of some early Tertiary mammals of North America. Bull. Neurol. Inst. N. Y., 1, pp. 430-505. Van Valen, L. 1966. Deltatheridia, a new order of mammals. Bull. Amer. Mus. Nat. Hist., 132( 1 ), pp. 1-126. Walker, E. P. 1964. Mammals of the world. Johns Hopkins Press, Baltimore, 3 vols.