L I E. R.ARY OF THL UN I VERSITY or ILLINOIS bb05 n REMOTE GEOLOGICAL SERIES OF FIELD MUSEUM OF NATURAL HISTORY folume VI Chicago, December 28, 1937 No. 19 SOME NOTOUNGULATE BRAINCASTS By Bryan Patterson Assistant Curator, Paleontology Results op the Marshall Field Paleontological Expeditions TO Argentina and Bolivia, 1922-27 In a recent paper of this series I published the results of a study of the middle ear structure of representative notoungulates (1936). This investigation was made in an attempt to discover heritage characters of possible phylogenetic and taxonomic importance. The results obtained were important enough to warrant testing them against evidence derived from another deep-seated structure — the brain. The results here presented are derived from a study of brain- casts of several important genera, and extend the excellent work of Simpson (1933a, b), who has described representative genera of several notoungulate families. Before describing the specimens, the technique employed in obtaining the casts may be briefly explained. With one exception these are artificial and were made by the use of a liquid rubber preparation. The extreme flexibility and ease of application of this substance render it ideal for the purpose. Liquid rubber is more convenient than glue or gelatine for this type of work, as it requires no heating and contains no water that would affect plaster joints in the specimen. No separator is needed. In dealing with friable bone, however, the surface to be cast should be thoroughly hardened with shellac or bakelite. The rubber penetrates every minute crack and crevice and is for this reason very hard to pull away from un- treated porous bone, and may even take away with it sections of such bone. Two to three coats of the liquid rubber are applied evenly to the bone with a brush. In dealing with a complete cranial cavity, the rubber may be poured in, rolled, and drained out. Each coat will dry within three or four hours. After application of the third or fourth coat, pulverized cork is dusted on to the wet rubber No. 400 273 L:-.,fary 274 Field Museum of Natural History — Geology, Vol. VI and allowed to dry with it. Successive coats of rubber and cork are applied until a thickness of a sixteenth to an eighth of an inch (2 to 4 mm.), depending upon the size of the specimen, has been reached. The cast is then allowed to dry for some twenty-four hours before removal. After being pulled, the cast is filled with plaster to preserve its shape while a mold of it is being made, preparatory to the making of permanent plaster casts. The durability of the rubber casts is not known. Some of those described here were prepared more than a year ago and appear to have suffered no deterioration. The liquid rubber solution used was obtained from a local firm, Devoe & Raynolds Company, 29 South Wabash Avenue, Chicago, at a cost of $1.55 per pint. I wish to express my thanks to Mr. Elmer S. Riggs for editing copy, to Mr. James H. Quinn for his skilful preparations, and to Mr. Frank H. Letl for introducing me to the liquid rubber technique. I am, as always, deeply indebted to Mr. Carl F. Gronemann for his painstaking work on the illustrations. Notohippidae Ameghino Rhynchippus Ameghino (fig. 75). Two braincasts of R. equinus Ameghino are available. The first is an excellent artificial cast taken from F.M. No. P13410 in which only the olfactory bulbs are imperfect; the second is a poor artificial cast taken from F.M. No. P13420, which has fairly good olfactory bulbs. Some notes are given below on a natural cast described by Professor Loomis under the name of Eutrachytherus spegazzinianus, but which unquestionably belongs in the Noto- hippidae. The Field Museum specimens were collected from the Deseado beds at Cabeza Blanca, Chubut, Argentina by the late Mr. J. B. Abbott. The total length of the skull of P13410 is 245 mm.; that of the braincast, exclusive of medulla, is approximately 93 (olfactory bulbs restored from P13420). The index is 38, and the ratio of olfactory , bulbs: cerebrum: cerebellum is about 3: 11: 5. Exposure of the olfactory bulbs, extent of overlapping of the cerebellum by the cerebrum, and degree of flexure are comparable to the conditions seen in Hegetotherium (Simpson 1933b). The rhinencephalon is well developed, but differs in details from that of most notoungulate braincasts hitherto described. In con- trast to the typotheres and in agreement with Rhyphodon (Simpson b>'z>0. O ^X \.G ^^ O&r^ -4 ri.U.s . NOTOUNGULATE BrAINCASTS 275 1933a) and Toxodon (Gervais 1872), the prominent olfactory bulbs are completely separated by a deep groove, which, however, does not affect the heavy deep peduncles, and is not continuous with the longitudinal scissure dorsally. The bulbs are laterally compressed and much deeper than wide, a marked difference from those of Rhyphodon. In this character there is an approach to Toxodon, ■v.p.gi>r ■sin. I. A.M. Laf. ce/r ven. sin m,IV.V",YT/ ^^^Y Ant/ier: vtiisia^--^- ■^- FiG. 75. Rhirnehippus equimis Ameghino. Dorsal, ventral, and left lateral views of braincast. F.M. No. P13410. Olfactory bulbs in part restored from F.M. No. P13420. For abbreviations see p. 301. XH- but in the Pampean form the bulbs are more widely separated. An interesting feature that has been noted is the presence of a foramen in the posterior part of the orbit, on the level of the optic foramen, which enters the olfactory fossa low on its lateral side. This foramen has been observed on both sides of P13410 and on the left side of P13420 (the right side is broken). From its position, it would seem possible that it transmitted the ethmoidal artery or arteries and perhaps the nasal nerve, if this nerve re-entered the cranial cavity in notoungulates. 276 Field Museum of Natural History — Geology, Vol. VI The olfactory tubercles are prominent, but not quite so conspicu- ous as in Hegetotherium. The pyriform lobes are large, but not visible in dorsal view. In outline and proportions they resemble those of Rhyphodon more closely than any form already described, and are not as expanded medially as those of Hegetotherium. They are notably shallower than the neopallium. In outline, the neopallium is roughly triangular, but it is relatively wider anteriorly than that of Hegetotherium. It definitely extends back beyond the pyriform lobes. The rhinal fissure is distinct anteriorly, but is concealed posteriorly by the large lateral cerebral venous sinus. No fossa Sylvii can be seen. The fissura Sylvii of the right side curves upward and backward almost in a semicircle; on the left side the posterior portion is lacking. In contrast to the typotheres, there appears to have been no suprasylvian sulcus. The sulcus lateralis is well defined anteriorly, poorly so posteriorly, and is completely interrupted in the middle of the cerebrum. It is somewhat curved, being closer to the longitudinal scissure at its center than at either extremity. Short crescent-shaped sulci are present in the coronal area, evidently homologous with the similarly situated depressions of Hegetotherium. The one on the right side is aligned antero-posteriorly with the horns of the crescent facing laterally, and is joined by the sulcus lateralis. The left sulcus is aligned transversely, with the horns facing anteriorly. The sulcus lateralis falls well short of it. The right coronal area extends further forward than the left. Between the coronal sulcus and the fissura rhinalis on the right side is an S-shaped, laterally-facing sulcus. On the left side there is only a small pit in this position. These depressions, and similarly situated ones in Nesodon and Adinotherium correspond, at least topographically, to the anterior portion of the sulcus in Proterotherium provisionally regarded by Simpson (1933b, p. 10) as ectosylvian. The longitudinal scissure is filled for its entire length by the external longitudinal sinus. The convolution pattern of Rhynchippus is about as definite as that of Hegetotherium, different in some features, and perhaps slightly more complex. It compares quite closely with that of Adinotherium to be described below. A striking feature of the cast at hand is the marked differ- ences between the sulci on the two sides of the neopallium. This feature is perhaps abnormal, but the perfect preservation of the skull indicates that it is not artificial. The hypophysis is nearly circular in outline, large, and very deep. It extends ventrally well below the pyriform lobes and the NOTOUNGULATE BRAINCASTS 277 fillings of the anterior lacerate foramina. Immediately behind it is a sharply defined concavity for the posterior clinoid process. Almost no structural details of the dorsal and occipital faces of the cerebellum can be seen. The dorsal exposure is narrow antero- posteriorly and wide transversely, equalling in width the anterior part of the neopallium. The occipital face slopes downward and backward at an angle of about 45°, and is gently convex transversely. Definite indication of lobes is lacking, and the postero-external angle is obscured by a thin lateral extension of the cast — the filling of a narrow gap between the postero-internal angle of the periotic and the supra- and exoccipital— which runs down to the foramen lacerum posterius. The lateral face is rough and rather complex. The internal auditory meatus is situated beneath the posterior extremity of the cerebrum. The ascending crus of the formatio vermicularis cannot be made out; the descending crus is stout. The "flocculus" is large and directly above the internal auditory meatus. Due to the fact that the dorsal portion of the medial face of the periotic slopes outward, the ascending crus and flocculus extend laterally over the internal auditory meatus. Anteriorly the periotic is produced into a stout medially expanded process which points downward and forward, forming a large cavity in the cast in front of the flocculus and meatus. In the region of the pons is a long gently convex area which fades away posteriorly at the level of the posterior lacerate foramina. The medulla is broader than deep and presents no structural peculi- arities. The nerve exits are all well preserved. The optic chiasma is stouter and much broader than in Hegetotherium, and is notably retracted above the fillings of the anterior lacerate foramina. The passages for III, IV, V'-^ and VI are very large and in the usual position. The foramen ovale (V^) is equally large, and, as described below, transmitted a large vein in addition to the nerve. The position of the internal auditory meatus (VII, VIII) with respect to the posterior extremity of the cerebrum, has been noted above. It is definitely on the lateral surface of the cast, a resemblance to Rhyphodon. IX, X, and XI were transmitted, as usual, through the foramen lacerum posterius. XII is single, of medium size, and further from the lacerate foramen than in Rhyphodon. The vascular sinuses are more conspicuous on this cast than on that of any notoungulate yet described. The large lateral cerebral 278 Field Museum of Natural History — Geology, Vol. VI venous sinus runs along the rhinal fissure, as in the typotheres but, in contrast to these forms, it passes to the outside of the skull through a foramen situated above the foramen opticum. This somewhat resembles the condition in Rhyphodon which, although lacking a lateral cerebral venous sinus, has a prominent vascular exit near to and above the optic and anterior lacerate foramina. The periotic of Rhynchippus is not united dorsally with the cranial roof, and in consequence a cavity exists above and lateral to it, which was not filled by any part of the brain, but which housed the very large postero-lateral venous sinus. The cast of the sinus is crescent-shaped with the horns of the crescent extending laterally and posteriorly. At its lateral extremity it receives two veins from the post-glenoid drainage. The lateral cerebral venous sinus enters anteriorly, and at the same point a sinus is given off ventrally which runs down to the foramen ovale. Two veins are received dorsally which enter the sinus through foramina opening externally along the squamoso-parietal suture. The posterior extremity leads to the mastoid foramen. Beneath the posterior extremity a poorly defined sinus leads ventrally, posterior to the periotic, to the foramen lacerum posterius where it undoubtedly passed to the internal jugular vein. The superior longitudinal sinus is visible along the entire length of the longitudinal scissure. At its posterior extremity it tends to bifurcate; this tendency is stronger on the Amherst cast, and is fully completed in the young Adinotherium described below. The two branches presumably joined the postero-lateral sinuses either directly or by joining a sinus running across the cerebellum, for there must have been one or more of these connecting the postero- lateral sinuses behind the blunt inconspicuous tentorium. One certainly connected, either directly across or via the postero-lateral sinuses (judging from the specimen, the latter alternative is more probable), the two sinuses which descend, posterior to the periotic, to the posterior lacerate foramina. This arch is unquestionably the homologue of the lateral (transverse) sinus of other mammals. This and the superior longitudinal are the only venous sinuses occur- ring in notoungulates that can be certainly homologized at present. To the others topographic names may be assigned. Simpson has already named the lateral cerebral and the postero-lateral. The sinus running down to the foramen ovale, I name anterior periotic. This appears to be the equivalent, if not the homologue, of the superior petrosal in man. The foramen, probably arterial, opening out of the olfactory fossa has been described above, NOTOUNGULATE BRAINCASTS 279 The brain of Rhynchippus appears to be approximately as pro- gressive as those of the typotheres, but is of a somewhat different type. The major distinctions of Rhynchippus are: 1. Olfactory bulbs completely separated. 2. Some differences in cerebral convolution pattern. 3. Pyriform lobes not as much expanded toward the mid-line. 4. Hypophysis much larger. 5. Lateral cerebral venous sinus passing anteriorly to the out- side of the skull. ^ 6. A venous sinus passing down to the foramen ovale. 7. Postero-lateral venous sinus widely open medially, due to failure of periotic to reach cranial roof. 8. Large anterior process of periotic. No. 8 is so far unknown in other notoungulates, 7 occurs in OldfieMthomasia (Simpson 1936, p. 27); the remaining distinctions are all shared with Nesodon and Adinotherium. Nos. 1 and 3, and in part 5, are seen in Rhyphodon; Nos. 1, 3, 5, and 6 are shared with Homalodotherium. Ameghino placed the Notohippidae in his order Hippoidea, but the characters of the braincast just described do not indicate any relationship to the Equidae. MEASUREMENTS IN MILLIMETERS Total length of skull 245 Length of brain, oblique, including medulla (olfactory bulbs restored from P13420) 109 Length of brain between verticals, excluding medulla (olfactory bulbs restored from P13420) 93 Length of olfactory bulbs (P13420) 17 Width across olfactory bulbs (P13420) 15 Length of cerebrum between verticals 54 Width of cerebrum 55 Length of cerebellum between verticals 25 Width of cerebellum across fiocculi 41 NOTES ON THE AMHERST CAST (Fig. 76) This specimen, a fairly complete natural cast with the dorsal surface of the cerebrum well preserved, was referred to Trachytherus {="Eutrachytherus") spegazzinianus by Professor Loomis (1914, pp. 77-80, fig. 47) on the basis of some associated fragments of the skull which are not now preserved. This identification is entirely 1 Not, perhaps, an important difference. Although the sinus apparently does not pass directly to the outside in typotheres, vascular foramina in the vicinity of the Sylvian fossa that may bear some relation to the sinus occur in several forms (Simpson 1936, p. 27). 280 Field Museum of Natural History — Geology, Vol. VI understandable, for parts of the skulls of Trachytherus are very similar to those of notohippids. I was able to show previously (1934, p. 126) that this cast was not referable to Trachytherus. The; series of casts prepared for this paper has now revealed its family position. The Amherst specimen was also collected at Cabeza Blanca, Chubut. The structure of the neopallium is sufficiently similar to that of the Rhynchippus just described that it will suffice to point out the Lar. ce.r. ven. s/'n. - R -I. ven. sin. Fig. 76. Notohippidae, gen. et sp. indet. Dorsal view of braincast. Amherst College Museum specimen. For abbreviations see p. 301. X Yi- distinctions. The fossa Sylvii appears to be faintly indicated. The , fissura Sylvii is not semicircular and runs into the sulcus lateralis; those of the two sides are symmetrical. The sulcus lateralis is con- tinuous on the right side, interrupted on the left; on both sides it runs forward into the coronal area. The left coronal area extends farther forward than the right. The longitudinal scissure is not obscured by the superior longitudinal sinus. Posteriorly the bifur- cation of this sinus can be seen. The posterior extremity of the neopallium slopes downward toward the tentorium much more sharply than in the cast described above. The cerebrum as a whole is longer and relatively narrower. I NOTOUNGULATE BrAINCASTS 281 Very little of the cerebellum is preserved, but what remains compares closely with that of Rhynchippus. The anterior process of the periotic is present and enclosed in the cast — a most important resemblance between the two specimens. Above this process the postero-lateral venous sinus is widely exposed. It is flatter than that of Rhynchippus, and does not show any indication that veins entered it dorsally. Professor Loomis interpreted this structure as the epitympanic sinus in the squamosal, but, as shown by many specimens, this cavity has no direct connection with the brain case in notoungulates. The lateral cerebral sinus is fairly well shown, but the sinuses anterior and posterior to the periotic are not preserved. The only nerve exits remaining on the cast are the fillings of the optic and anterior lacerate foramina, and the internal auditory meatus; these are identical with those of Rhynchippus. Every feature of this specimen, and particularly the anterior process of the periotic and the open posterolateral venous sinus, indicates that it should be referred to the Notohippidae. Generic identification is not possible at present. Although the extent of individual variation in Rhynchippus is not known, and the distor- tion, if any, undergone by the skull from which the Amherst cast came will never be known, the specimen can hardly be referred to this genus. Several notohippids — Morphippus, Argyrohippus, Core- sodon, etc. — to which it might be referable occur in the Deseado. Until more information is available, the specimen must be referred to as Notohippidae, gen. et sp. indet. Toxodontidae Nesodon Owen (fig. 77). An artificial braincast has been taken from F.M. No. P13076, referred to N. imbricatus Owen, and collected by Mr. G. F. Stern- berg from the Santa Cruz beds on the Atlantic coast eight miles South of the Rio Coyle. The specimen has suffered a little lateral compressure, which seems to have slightly increased the height and decreased the width of the cast. The amount of distortion cannot be gauged accurately and has, therefore, not been corrected in the drawings. The total length of the skull is 473 mm., that of the braincast, enclusive of medulla, 118. The index is 25, and the ratio of olfactory bulbs: cerebrum: cerebellum is approximately 3: 13: 7. Taking into account the large size of the skull and the relatively shorter olfactory bulbs — factors tending to give a low index — the brain of 282 Field Museum of Natural History — Geology, Vol. VI Nesodon appears to be somewhat more advanced than that of Rhynchippus, and less macrosmatic. As stated previously (Riggs and Patterson 1935, p. 191), the toxodontid sl-oill is cyptocephaUc, i.e., the basifacial axis is bent down upon the basicranial. The occiput is pitched forward. As a result of this bending down of the facial region and of the inclination of the occiput, the cerebrum is inclined upward and quite strongly arched, much of the neopallium is dorsal to the cerebellum, and the dorsal exposure of the cerebellum slopes downward and backward. The olfactory bulbs are short in comparison with those heretofore described, deep, and narrow transversely. As in Rhynchippus they are separated by a groove which neither splits the peduncles nor joins the longitudinal scissure. Resemblance to Toxodon is close, with the exception that in the Pampean genus the bulbs are farther apart. The olfactory peduncles are large and long. The pyriform lobes are relatively small; structurally they are very similar to those of Rhynchippus. They are lateral in position, invisible in dorsal view, and little conspicuous in ventral view. This is perhaps a little accentuated by the slight distortion of the skull, but it is certainly a natural feature of the brain and is related to the upward inclination of the cerebrum and the large size of the hypophysis. The neopallium is similar in outline to that of Toxodon. It is somewhat wider anteriorly than that of Rhynchippus and notably more so than those of the typotheres. The anterior portion of the fissura rhinalis is very distinct, but posteriorly it is, as usual, concealed by the lateral cerebral venous sinus. The fossa Sylvii is in the usual position, but very poorly defined. The fissura Sylvii is very distinct, as in Toxodon, and rather undulating. The short sulcus situated dorsal to the anterior part of the rhinal fissure is longer than in Rhynchippus, extending posteriorly to the sylvian fissure. The longitudinal scissure is filled by the superior longitudinal sinus which is prominent anteriorly, as in Toxodon, and medianly. No other sulci can be seen on the dorsal surface of the cerebrum, and the same seems to be true of Toxodon, to judge from Gervais' figure. The cerebrum of the young Adinotherium ovinum described below shows that these conditions are due to lack of impression of the sulci on the roof of the brain case, and not to extreme simplicity of pattern in the family. The hypophysis is enormous and very deep, deeper even than in Rhynchippus. It extends ventrally far below the pyriform lobes, well below the fillings of the anterior lacerate foramina, and forward ^ .s. Cl.pf - IX,X,XI I. A.M. Fig. 77. Neaodon imbrieatus Owen. Dorsal, ventral, and right lateral yiews of braincaat. P.M. No. P13076. The olfactory btilbs may not be quite as divergent as represented. For abbreviations aee p. 301. X ^i. 283 284 Field Museum of Natural History — Geology, Vol. VI between these fillings. Between the anterior extremity of the hypophy- sis and the fillings of the anterior lacerate foramina are two deep concavities in the cast caused by bony processes extending back from the sides of the optic chiasma. A foramen, probably vascular, leads forward from the mid-line of the ventral extremity of the hypophysis. The posterior clinoid process is elongate and flanked on either side by depressions which appear on the cast as ridges. The tentorium of Nesodon is a rather narrow blunt process situated on the mid-line of the skull and not extending completely across the roof of the cranial cavity. It forms a large depression in the cast, on either side of which the dorsal surfaces of the cerebrum and cerebellum appear as though confluent. The cerebellum is narrower than the anterior portion of the cerebrum. The occipital face slopes downward and backward at about the same angle as that of Rhynchippus. Due to the upward inclination of the cerebrum, the dorsal face is not horizontal, as it is in other notoungulates (Toxodon excepted) hitherto described, but also slopes downward and backward, although at a more gentle angle than the occipital face. As in Rhynchippus, no traces of structure are apparent. Also as in this genus, the periotic is not firmly united postero- internally with the supra- and exoccipital. A narrow gap is thus left in the cranial wall which was probably not occupied by any part of the brain, but which is represented on the cast by a thin postero- lateral extension. In contrast to Rhynchippus, the internal face of the periotic is smooth and vertical. The formatio vermicularis is not apparent, and the. flocculus is indicated only by a very faint convexity above the internal auditory meatus. The pons is very similar to that of Rhynchippus but occupies a rather more posterior position in relation to the periotic. The medulla is somewhat longer, and shallower dorso-ventrally. All nerve exits are preserved. The optic chiasma is situated more in advance of and above the anterior lacerate foramina than in Rhynchippus, due to the upward inclination of the cerebrum. The passages for III, IV, Y^-^, and VI are somewhat higher on the sides of the cast and are more widely separated than in the last mentioned genus. They are much more widely separated than in any known typothere or in Notostylops, and resemble those of Phenacodus (Simpson 1933a, figs. 1-2) in this respect. The filling of the foramen ovale (V^ and the anterior periotic venous sinus) is narrow antero-posteriorly and extends out for some distance from the body of the cast. The internal auditory meatus (VII, VIII) is NOTOUNGULATE BRAINCASTS 285 situated high on the side of the cast and nearly in the center of the periotic face of the cerebellum. It extends far into the periotic before the nerves separate. Above the foramen lacerum posterius (IX, X, XI) is a large irregularly shaped recess to be described below. The exit for XII occupies the same position with relation to the filling of the lacerate foramen as does that of Rhynchippus. The vascular sinuses present some differences from those of Rhynchippus, but the general similarity is very close. The lateral cerebral venous sinus is in its usual position on the posterior half of the rhinal fissure, and opens anteriorly above the filling of the foramen lacerum anterius. In contrast to Rhynchippus, it does not join the anterior periotic venous sinus on the cast, the two sinuses evidently uniting within the cranial wall. A conspicuous anterior vascular opening is present in the anterior half of the rhinal fissure, a structure not present in Rhynchippus. The periotic of Nesodon differs from that of the last named genus in being closely approxi- mated to, although not fused with, the cranial roof. As a result, the postero-lateral venous sinus is not open internally. However, due to the imperfect union of periotic and cranial roof, a thin supero- lateral extension of the cast projects for a short distance between the two. This space was evidently not occupied by any part of the brain, but it seems to have provided passage for a sinus or sinuses (fig. 77). A peculiar feature of the cast is the greatly expanded recess above the foramen lacerum posterius, a structure not en- countered in the other forms studied, but clearly shown in Roth's excellent sagittal section figures of Toxodon burmeisteri (1898, pis. 5, 6). This recess was evidently filled in life by nerves IX, X, and XI, the internal jugular vein, and part of the lateral sinus. The last named may have been unusually large. In addition, a vascular tract, possibly part of the inferior periotic sinus, runs backward and slightly downward from the upper part of the recess to the hypo- glossal canal. The superior longitudinal sinus has been mentioned in the description of the neopallium. Many small blood vessel impressions may be seen on the dorsal surface of the cerebrum. Several of these unite to form two moderately large vascular tracts which run posteriorly, one on either side of and close to the superior longitudinal sinus. The fundamental similarities of the braincast of Nesodon to that of Rhynchippus have been given above (p. 279) The chief characters in which Nesodon differs from Rhynchippus are as follows: 1. Smaller olfactory bulbs. 286 Field Museum of Natural History— Geology, Vol. VI 2. Cerebrum higher, more arched and indined upward. 3. Dorsal exposure of cerebellum sloping downward. 4. A vascular opening on the anterior portion of the rhinal fissure. 5. Lateral cerebral and anterior periotic sinus meeting outside the cerebral cavity. 6. A large recess dorsal to the foramen lacerum posterius. Of these, 2 and 3 are to be correlated with the cyptocephalic skull, and may be regarded as habitus distinctions. No. 5 is caused by the periotic reaching the cranial roof and shutting off the postero- lateral venous sinus from the cranial cavity. As stated above, the open condition of the sinus is a peculiarity of Rhynchippus, and may well be a family character. Nos. 1, 2, 3, and 6 will probably prove to be family characters of the Toxodontidae. These differ- ences between Rhynchippus and Nesadon are of less importance than the resemblances common to both, which are listed on p. 279. They do not oppose the conclusion that the two genera are funda- mentally similar in brain as well as in ear structure. An important fact brought out by this investigation is the revela- tion of the very close resemblance, almost amounting to identity, between Nesodon and Toxodon in endocranial structure. The shape of the cranial cavity of Toxodon, as shown by Roth's admirable figures (1898, pis. 5, 6), reveals that the cerebrum was inclined dorsally and arched, more so indeed than that of Nesodon; that the dorsal exposure of the cerebellum sloped downward; that a large recess was present above the foramen lacerum posterius; and that the olfactory bulbs were proportionately the same size as in Nesodon. In short, the evidence of endocranial anatomy is decidedly opposed to recognition of a distinct family, Nesodontidae. MEASUREMENTS IN MILLIMETERS Total length of skull 473 Length of brain, oblique, including medulla 150 Length of brain between verticals, excluding medulla 118 Length of olfactory bulbs 14 Length of cerebrum between verticals 69 Width of cerebrum 58 Length of cerebellum between verticals 35 Width of cerebellum above I. A. M 31 Adinotherium Ameghino (fig. 78). An incomplete cast of the cerebrum was obtained from F.M. No. P13107, a young individual of A. ovinum (Owen) which has yielded valuable information concerning details of the ear structure (Patterson 1936, p. 214, fig. 50). It was collected by Mr. Abbott from the Santa Cruz beds five miles south of Coy Inlet. The right NOTOUNGULATE BrAINCASTS 287 side of the cast is well preserved; the left is considerably distorted. The importance of the specimen lies in the fact that it reveals almost the entire cerebral convolution pattern which is not well shown in Gervais' specimen of Toxodon, nor in the cast of Nesodon described above. The outline of the neopallium is essentially the same as in Nesodon. The rhinal fissure is obscured anteriorly by crushing, and Ldt, cer. ven. s/'n. Fig. 78. AdinoQieTiwrn, ovinum (Owen). Dorsal view of cast of cerebrum of young individual. F.M. No. P13107. For abbreviations see p. 301. X}. posteriorly by the lateral cerebral venous sinus. As in Rhynchippits, the fossa Sylvii cannot be seen. The fissura lateralis is similar in general to that of Rhynchippus, but is more definite, more curved, and more erratic in its course. It does not extend to the coronal area, in which there is, on the right side at least, a slight depression. The sulcus situated between the anterior portions of the fissurae lateralis and rhinalis is practically identical with that of Nesodon. The longitudinal scissure is visible anteriorly, but is completely con- cealed posteriorly by sinuses. The close agreement of this convolution pattern with that of Rhynchippus strengthens the belief expressed above concerning the close relationships of the Toxodontidae and Notohippidae. 288 Field Museum of Natural History — Geology, Vol. VI The cast shows certain of the venous sinuses and their tributaries extremely well. The lateral cerebral is in its usual position along the posterior portion of the rhinal fissure. Near its anterior exit it receives two dorsal tributaries. The anterior of these, which is almost as large as the sinus, divides into four small branches at a point seven mm. above the junction. Two of these branches extend forward, one extends upward, and the fourth and longest extends upward and backward. The posterior tributary is small and divides almost immediately into two branches, one of which roughly parallels the sinus, the other the long branch of the anterior tributary. A small sinus with an anterior exit is situated beneath the small sulcus between the fissurae lateralis and rhinalis. This may be the homo- logue of the vascular opening on the anterior half of the rhinal fissure in Nesodon, but there can be no certainty on this point due to the obscuring by crushing of this fissure on the cast at hand. The superior longitudinal sinus is more complex (or perhaps more detail in preserved) on this specimen than on any other yet described. It is apparent on the posterior half of the cerebrum only, becoming indistinct near the center of the longitudinal scissure. At this point it receives a large tributary on the left side which proceeds for a short distance antero-externally before fading from view. This tributary in turn receives a small but sharply defined branch which runs postero-externally for some seven mm. before suddenly ending. Some six mm. posterior to the point of entry of the left tributary, the sinus receives a tributary on the right side. This extends for- ward, then loops backward, and then loops forward again to disappear on the fissura lateralis. On the posterior third of the longitudinal scissure, the sinus divides into right and left portions which curve externally around the median parts of the posterior extremities of the cerebral hemispheres. The right is lower than the left, and receives a weW defined tributary which parallels it for its entire length and passes out of the cranial cavity into the cranial roof. Homalodotheriidae Ameghino Homalodotherium Flower (fig. 79). The partial skeleton of H. cunninghami in the Field Museum collections, No. P13092, was recently mounted by Mr. Quinn, and advantage was taken of further preparation of the skull to obtain an excellent braincast. This fine specimen, collected by Mr. Riggs in the vicinity of Cape Fairweather, Santa Cruz, has already been .S.3. - -O.B. Anh per: ven. s/n. Fig. 79. Hs/n 3.S. , Lat- cer. vizn. sm. P.L.. Fig. 80. Typotheriopsis internum (Ameghino). Dorsal and left lateral views of incomplete brain- cast. F.M. No. P14420. For abbreviations see p. 301. X %. not a conservative, surviving side branch, that a certain amount of brain evolution has taken place in the family since Oligocene times. The olfactory bulbs are fully exposed, the midbrain was certainly covered, but the cerebrum has barely begun to overlap the cerebellum. The flexure is definitely slighter than in Hegetotherium, and agrees NOTOUNGULATE BRAINCASTS 295 closely with that seen in Trachytherus. The brain appears to have been shallower than that of Hegetotherium, as is also true of Typo- therium, to judge from a section of the cranium given by Gervais (1867, pi. 23, fig. 4). The ratio of cerebrum to cerebellum is approxi- mately 2:1, as in Trachytherus; restoring the olfactory bulbs, the ratio was probably about 4:11:5, slightly less advanced than that of Hegetotherium. The rhinencephalon is as well developed as in other known typo- theres. The olfactory bulbs are united in the parts preserved, and were probably separated only at their extremities. The peduncles, and probably the bulbs also, are narrower than those of Hegeto- therium; they are also somewhat longer.^ In Protypotherium the peduncles agree in length with those of Typotheriopsis, but are some- what wider, although narrower than those of Hegetotherium. The olfactory tubercles are not preserved. The pyriform lobes are largely destroyed, but enough remains to show that they were similar to those of Trachytherus and Hegetotherium. The outline of the neopallium is similar to that of Protypotherium, but is slightly narrower posteriorly than in this genus or in Hegeto- therium. The rhinal fissure is concealed by the postero-lateral cerebral venous sinus. The fossa Sylvii has been somewhat obscured by slight distortion but appears to have been small. The fissura Sylvii is well marked and runs upward and backward. The sulcus lateralis is rather poorly defined and, in contrast to Hegetotherium, does not have curving ends. As in this genus, there are pits in the coronal area anterior to the sulcus lateralis. The longitudinal scissure is very distinct anteriorly; posteriorly the superior longi- tudinal sinus is plainly visible, and broad. The suprasylvian sulcus is very slight and, as in Protypotherium, has no anterior part. The sulci as a whole are not as well defined as those of Hegetotherium, resembling those of Protypotherium in this respect. They are, how- ever, much plainer than in Gervais' specimen of Typotherium, as figured (1872, pi. 21, fig. 11). The cerebellum, as a whole, is not as low relative to the cerebrum as is that of Hegetotherium. The occipital exposure is much more sloping than in the last named genus; there is no apparent separation between the dorsal and occipital faces. Very little detail is apparent; the occipital face shows a broad posteromedian lobule similar to that of Hegetotherium. The lateral surfaces of this face are obscured 1 If Typotheriopsis is a reliable criterion, the restored olfactory peduncles in the published figure of Trachytherus (Patterson 1934, fig. 28) are too wide. 296 Field Museum of Natural History— Geology, Vol. VI by the sinuses described below. The lateral face is, as a whole, decidedly convex. The formatio vermicularis is prominent, the descending crus and flocculus being more definite than in Hegeto- therium. The flocculus is directly above the internal auditory meatus, as in Trachytherus. The periotics of Typotheriopsis are closer to the midline of the skull than in Hegetotherium and Protypo- therium, and the cerebellum is consequently narrower anteriorly. The internal auditory meatus is the only nerve exit preserved. This opening and the flocculus are rather closer to the cerebrum than are those of Trachytherus, but, in contrast to Hegetotherium, are still not beneath the posterior end of the cerebrum. The posterolateral venous foramen leading from the cerebral cavity is not as large as in Hegetotherium, nor is it situated as high nor as far back. The venous sinus running from it turns down quite sharply after leaving the foramen, and then runs forward along the fissura rhinalis. No traces of minor veins leaving the foramen or entering the sinus can be detected either on the cast or on the inside of the brain case. On the posterolateral angle of the cerebellum is a conspicuous, dorso-ventrally aligned sinus which emerges from a foramen situated slightly above the level of the flocculus. Ventrally it continues in the direction of the foramen lacerum posterius (not preserved on the cast). From its relations, this sinus would appear to be the homologue of that regarded as the lateral in the casts described above. At the level of the internal auditory meatus it gives off a branch which runs posteriorly along the side of the medulla oblongata. This branch receives a somewhat smaller tributary which, at first appearing near the junction of the medulla with the occipital face of the cerebellum, runs parallel to and slightly above it for a short distance. The cast is not complete enough to follow this branch for its entire course. Judging from other skulls it appears to end in the vicinity of the condylar foramen. It is probably homologous, therefore, with the vascular tract which runs back toward this foramen in Nesodon (see p. 285). The periotic of Typotheriopsis is closely united with the roof and sides of the cranial cavity. The brain of Typotheriopsis is in general very similar to those of other typotheres that have been described. It appears to be some- what more primitive than that of Hegetotherium, although from a considerably more recent horizon, in a few features such as slightly lesser depth, lesser degree of flexure, and more sloping occipital face of cerebellum. In all parts in which comparison is possible, the NOTOUNGULATE BRAINCASTS 297 braincast agrees closely with that of Trachytherus, thus further confirming the reference of this genus to the Tj-potheriidae. I have previously stated (1936, p. 224) that in some characters the Intera- theriidae and Typotheriidae were closer to each other than to the Hegetotheriidae. Simpson has listed (1933b, pp. 6-7) the few characters in which the braincast of Protypotherium differs from that of Hegetotherium. Typotheriopsis resembles Protypotherium in these characters, with the exception of 3, which is not a structural differ- ence. Sagittal sections of Protypotherium and Interatherium show that these genera agree with Typotheriopsis and differ from Hegeto- therium in having a shallower brain, a slighter degree of flexure, and a higher cerebellum. The observed differences between the Hegetotheriidae on the one hand and the Typotheriidae-Intera- theriidae on the other are not striking, however, and from the avail- able evidence it would appear that the Typotheria, as currently recognized, form a compact uniform group in brain structure. The possibility that this observed uniformity does not indicate taxonomic unity is discussed below. MEASUREMENTS IN MILLIMETERS Total length of skull (estimated, from P14477) 265 Length of brain between verticals, excluding medulla (olfactory bulbs estimated) 95 Length of cerebrum between verticals 55 Width of cerebrum 52 Length of cerebellum between verticals 21 Width of cerebellum 28 CONCLUSIONS Study of the middle ear structure in the Notoungulata (Patterson 1936) has indicated that, as far as this highly important region is concerned, the families Toxodontidae, Leontiniidae, and Noto- hippidae (the suborder Toxodonta of various authors) form a natural group. It has also been shown that Homalodotherium, generally regarded as the typical representative of the suborder Entelonychia, is clearly referable to this group on the evidence of middle ear struc- ture. The families Typotheriidae and Interatheriidae have been found to possess several important characters in common, characters in which both of them differ from the Hegetotheriidae. The validity of the Entelonychia, and the unity of the suborder Typotheria, as currently recognized, have thus been called into question. The bearing of evidence derived from the braincasts described above on these problems may now be reviewed. 298 Field Museum of Natural History — Geology, Vol. VI The close agreement in brain structure of the Notohippidae and Toxodontidae, as exemplified by Rhynchippus, Nesodon, and Adinotherium, and their common divergence from the typotherian brain type have already been discussed, and require no further comment. Homalodotherium agrees with these genera in the majority of the characters in which they differ from the tjrpotheres (p. 279), namely, separation of olfactory bulbs, position and expansion of pyriform lobes, and relations of lateral cerebral and anterior periotic venous sinuses. The hypophysis is not expanded ventrally in Homalodotherium, but it may be noted that the space between the fillings of the anterior lacerate foramina is fully as large as in Nesodon. The non-expanded condition may perhaps be correlated with the large size of the frontal sinus. The cerebral convolution pattern of Homalodotherium is unfortunately unknown, and may or may not have been similar to the toxodontid-notohippid pattern. Rhy- phodon displays some differences from this pattern, but it has been shown above that this genus is aberrant in certain features. There- fore, it may not be typical of the homalodotherid pattern. The notable characters in which Homalodotherium differs from Nesodon and Rhynchippus — depressed olfactory bulbs and anterior portion of cerebrum, and non-retracted optic nerves — are correlated with the early development of a large frontal sinus in the family, and hence do not, in my opinion, offset the positive resemblances that have been noted between the three casts. I believe, therefore, that the evidence obtained from this study supports the conclusion reached from middle ear structure that "the homalodotherids stand in about the same relation to the Toxodonta as the chalicotherids do to the less aberrant perissodactyls." This conclusion carries with it the implication that the suborder Entelonychia should no longer be recognized, and that its component families Homalodo- theriidae and Isotemnidae should be transferred to the Toxodonta. Simpson has recently shown (1936b) that the peculiar skeletal modi- fication of the homalodotherids has been acquired since Musters time, the earlier members of the family having skeletons of more generalized type. This seems to be in keeping with the subordinal assignment made here. It may be necessary to separate the homalo- dotheres from the remaining toxodonts either as an infraorder or as a superfamily, but this question is beyond the scope of the present study. If the views expressed here are correct, the braincasts of Rhyn- chippus, Nesodon, and Homalodotherium should be structurally NOTOUNGULATE BRAINCASTS 299 derivable from a common stem tjrpe. Recognition of such a type will be greatly impeded by the major gap in the South American record that exists between the Musters of middle or upper Eocene age and the Deseado of middle or upper Oligocene age. The im- portance of this unrecorded interval is admirably demonstrated by the homalodotherids which, on present evidence, appear to have undergone practically their entire profound skeletal modification during it. As stated above, Thomashuxleya and Periphragnis of the Casamayor and Musters are generalized notoungulates in skeletal characters (Simpson 1936 b, p. 12), whereas Asmodeus of the Deseado is about as fully specialized as Homalodotherium (Loomis 1914, figs. 94-99, and original specimens). The same probably holds true for other notoungulate groups, all of which undergo only relatively minor modifications in post-Deseado time. In studying the brain- casts of Rhynchippus, Nesodon, and Homalodotherium, we are dealing with fully specialized representatives of three divergent families, and have no present opportunity adequately to trace the early stages of the divergence. In spite of these difficulties, I am, nevertheless, inclined to believe that the braincast of a generalized toxodont such as the Casamayor isotemnid PUurostylodon, should present a fairly close approximation to the hypothetical prototype. The brain- cast of Rhypkodon probably approaches this ideal, but it is definitely homalodotherid in some features and seemingly aberrant in others, and accordingly does not fulfill the necessary requirements. The known typotherian braincasts are all of similar type, whereas the middle ear structure of these animals indicates a division into two groups. The known toxodont braincasts exhibit a considerable degree of divergence, whereas the middle ear structure is very uni- form throughout. This apparent contradiction might be considered as invalidating the conclusions arrived at above. I do not believe that this is the case, however. The brain, being the dominant organ of the body, is more liable than the middle ear to modification in response to varying modes of life. The Hegetotheriidae, Intera- theriidae, and Typotheriidae have in common a remarkably rodent- like habitus. Simpson (1933a, b) has demonstrated that the braincast of Notostylops of the Notioprogonia, also a rodent-like form, is fundamentally similar to the typotherian type. It seems entirely possible, therefore, that the Hegetotheriidae on the one hand and the Typotheriidae-Interatheriidae on the other, may have descended independently from notioprogonian ancestors, and that the close similarity they display in braincast structure may be in 300 Field Museum of Natural History — Geology, Vol. VI part convergent, due to the acquisition of almost identical habitus, and in part common heritage. The proper taxonomic position of the Hegetotheriidae cannot be decided, however, on the evidence at present available. REFERENCES Cabrera, C. and Kraglievich, L. 1931. Diagnosis previas de los ungulados f6siles del Arroyo Chasico. Notas prelim. Mus. La Plata, 1, 107-113. Gervais, p. 1867-9. Recherches sur differents groupes de mammiferes fossiles de I'Ameri- que meridionale. Zool. Pal6ont. gen^rales, 129-147, 252-254, pis. 22-25, 27. 1872. Memoire sur les formes cerebrales propres a differents groupes de mammiferes. Journ. Zool., 1, 425-469, pis. 20-23. Gregory, W. K. 1936. On the phylogenetic relationships of the giant panda (Ailuropoda) to other arctoid carnivora. Amer. Mus. Novit. No. 878, 1-29, figs. 1-22. Kraglievich, L. 1934. La antigiiedad pliocena de las faunas de Monte Hermoso y Chapadmalal, deducidas de su comparacion con las que precedieron y sucedieron. Imprenta "El Siglo Illustrado," Montevideo. Pp. 1-136. A posthumous work, pp. 1-16 being devoted to a brief obituary notice and a bibliography. Loomis, F, B. 1914. The Deseado formation of Patagonia. Amherst, 1-232, frontispiece, text-figs. 1-160. Patterson, B. 1934. Trachytherus a typotherid from the Deseado beds of Patagonia. Geol. Ser. Field Mus. Nat. Hist., 6, 119-139, text-figs. 24-28. 1936. The internal structure of the ear in some notoungulates. Ibid., 199-227, text-figs. 44-54. Riggs, E. S. and Patterson, B. 1935. Description of some notoungulates from the Casamayor (Notostylops) beds of Patagonia. Proc. Amer. Phil. Soc, 75, 163-215, text-figs. 1-4, pis. 1-5. Roth, S. 1898. Catalogo de los mamiferos fosiles conservados en el Museo de La Plata. Grupo Ungulata, orden Toxodontia. Rev. Mus. La Plata, 8, 33-160, text-figs. 1-81, pis. 1-8. Simpson, G. G. 1933a. Brain casts of Phenacodus, Notostylops, and Rhyphodon. Amer. Mus. Novit. No. 622, 1-19, text-figs. 1-3. 1933b. Brain casts of two typotheres and a litoptern. Ibid., no. 629, 1-18, text-figs. 1-3. 1936a. Structure of a primitive notoungulate cranium. Ibid., No. 824, 1-31, text-figs. 1-11. 1936b. Skeletal remains and restoration of Eocene Entelonychia from Pata- gonia. Ibid., No. 826, 1-12, text-figs. 1-2. NOTOUNGULATE BRAINCASTS 301 ABBREVIATIONS ON FIGURES II Optic nerve III, IV, V'-*, VI Filling of foramen lacerum anterius, the common exit of these nerves V* Mandibular nerve IX, X, XI Filling of foramen lacerum posterius, the common exit of these nerves XII Condylar foramen (hypoglossal canal) Ant. per. ven. sin. Anterior periotic venous sinus CI. pr. Depression caused by posterior clinoid process F. L. "Flocculus" F. R. Fissura rhinalis F. S. Fossa Sylvii H. Filling of fossa hypophyseos I. A. M. Internal auditory meatus (nerves VII, VIII) Lat. cer. ven. sin. Lateral cerebral venous sinus Lat. sin. Lateral (transverse) sinus O. B. Olfactory bulb O. T. Olfactory tubercle P. L. Pyriform lobe P.-l. [post-lat.] ven. sin. Posterolateral venous sinus Rec. pr. per. Recess formed by anterior process of periotic S. L. Sulcus lateralis S. S. Fissura Sylvii S. S. S. Suprasylvian sulcus Sin. lat. Lateral (transverse) sinus V. Vascular opening V. p.gl. f. Veins passing to postglenoid drainage V. sq.-par. sut. Veins passing to foramina along the squamoso- parietal suture Ven. sin. Venous sinus