AMERICAN MUSEUM NOVITATESI Published by Number 1207 the AMERICAN MUSEUM of NATURAL HISTORY November 12, 1942 New York City

AMERICAN MUSEUM NOVITATESI Published by Number 1207 THE AMERICAN MUSEUM OF NATURAL HISTORY November 12, 1942 New York City

NOTES ON THE LESSER ONE-HORNED RHINOCEROS, RHINOCEROS SONDAICUS

2. THE POSITION OF RHINOCEROS SONDAICUS IN THE PHYLOGENY OF THE GENUS RHINOCEROS By EDWIN H. COLBERT

Although there is a fairly extensive tions as to its relationships to some of the literature on the so-called Javan rhinoceros, Asiatic Pleistocene forms became at once Rhinoceros sondaicus, this form has been apparent, thereby affording a much clearer generally neglected in discussions having picture of the phylogenetic development of to do with the phylogenetic development the genus Rhinoceros than had previ- of the Rhinocerotidae. It is true, of ously been envisioned by the writer. Con- course, that various authors have recog- sequently it has been thought advantage- nized Rhinoceros sondaicus as an animal less ous to discuss briefly the evolution of "specialized" or less "modified" than the Rhinoceros (using the term in a strict sense) Indian rhinoceros, Rhinoceros unicornis in an effort to show how Rhinoceros (Osborn, 1898; Matthew, 1931), but be- sondaicus may be related to other species of yond such general statements as to its the genus. phylogenetic relationships there has been In the first place, it might be well to indi- little or no effort made to discuss in a de- cate the limits of the genus Rhinoceros. tailed way the affinities of this extraor- This genus has been used, particularly by dinarily rare and interesting mammal. palaeontologists, to include a great number Perhaps this neglect of Rhinoceros sondaicus of species ranging in age from the Miocene in the more general discussions of phy- through the Pleistocene and into Recent logeny within the family Rhinocerotidae has times. Indeed, during the early history been due, to a considerable extent, to the of vertebrate paleontology it was the prac- comparative rarity of the species in muse- tice to designate almost every fossil rhino- ums, or at least to the lack of acquaint- ceros'6f post-Oligocene age as Rhinoceros, ance with this animal by palaeontologists. since in those days students of fossils For it is only through a palaeontological naturally were but little concerned with the background, particularly a knowledge of criticalt limitations for genera and species the late Tertiary and Quaternary rhinocer- that have developed with the refinement of oses of Asia, that the true position of the science. Rhinoceros sondaicus in its relationships to Obviously this is wrong and has been so other rhinoceroses can be' completely recognized with the increasingly detailed appreciated. studies that have been made in later years It was the good fortune of the present on fossil vertebrates. Consequently at the author to be engaged in a study of the ex- present time most of the fossil species tinct rhinoceroses from the Pleistocene of formerly designated as Rhinoceros have China-as part of a general study of the been allocated to other genera, and the Pleistocene mammalian fauna from the genus has thus become strictly limited ac- limestone fissures of Szechwan-when the cording to the evidence of its anatomical skull of Rhinoceros sondaicus in the Ameri- characters, which, as far as the osteology of can Museum came to light. When this the skull and the dentition are. concerned, recent skull was examined, certain implica- are as follows: 2 AMERICAN MUSEUM NOVITATES [No. 1207

1. Expansion of the nasal bones into a Recent of the Sundarbans, eastern Bengal, boss or eminence, upon which is borne the Assam, Burma, Malay Peninsula, Sumatra single nasal horn. and Java. 2. Incisors present, and of large size. 3. Skull short, with occipital plane in- Rhinoceros sivalensis Falconer and Cautley, clined forward. 1847. 4. Auditory meatus closed inferiorly by Synonym: R. palaeindicus. fusion of the post-tympanic and the post- Pleistocene of the Siwalik Hills of India. glenoid processes. Rhinoceros sinensis Owen, 1870. 5. Cheek teeth sub-hypsodont. Synonyms: R. plicidens, R. simplicidens. Upon the basis of the above limitations, Pleistocene of southwestern China. the genus Rhinoceros includes four good It might be well at this place to indicate species, two of which are of recent age, and the distinctions in the skull, jaw and denti- two of which are extinct. There are other tion between the two recent species, fossil species of doubtful validity which Rhinoceros unicornis and Rhinoceros son- need not be considered at this place. The daicus, since many of these distinctions are four species with which we are concerned of considerable importance in the following are: discussion of the phylogenetic position of Rhinoceros unicornis Linnaeus, 1758. Type each species in relation to the other. of genus. Moreover, by calling attention to these Synonyms: R. indicus, R. asiaticus,. R. osteological differences between the two stenocephalus. modern species it may be that some aid will Pleistocene and Recent of India. Now be afforded to workers in the future who very limited in range. may have occasion to examine skulls and teeth of Rhinoceros. Rhinoceros sondaicus Desmarest, 1822. Flower (1876) published a paper in which Synonyms: R. javanicus, R. inermis, R. he described many of the distinguishing nasalis, R. floweri. traits between the two living species of Possibly Pleistocene (of Borneo) and Rhinoceros, while Osborn in his monograph

SKULL, MANDIBLE AND DENTITION Rhinoceros unicornis Rhinoceros sondaicus 1. Large and robust. 1. Smaller and lighter than R. unicornis. 2. Nasals expanded into large, rounded horn 2. Less expansion in the nasals; horn boss boss. pointed rather than rounded, and very small in female. 3. Ascending ramus very high. 3. Ascending ramus not extremely heightened. 4. Occipital surface high and narrow. Skull 4. Occipital surface comparatively low and deep. broad. Skull comparatively shallow. 5. Deep "saddle" in profile of skull, between 5. Rather shallow saddle in cranial profile. nasals and occipital vertex. 6. Zygomatic arch rounded at posterior ter- 6. Zygomatic arch angular at posterior termination. mination. 7. Posterior margin of palate concave, or with 7. Posterior margin of palate with median small median projection. projection. 8. Mesopterygoid fossa, basisphenoid and basi- 8. Mesopterygoid fossa, basisphenoid and basi- occipital bones narrow. occipital bones comparatively broad. 9. Pterygoids compressed and grooved. 9. Pterygoids flattened and laterally expanded. 10. Vomer thick and united to sides of pterygoid 10. Vomer thin, lamelliform, pointed and free. processes. 11. Premaxillaries broad. 11. Premaxillaries relatively narrow. 12. Teeth strongly sub-hypsodont. 12. Teeth less hypsodont than in R. unicornis. 13. Ectoloph of cheek teeth rather flat. 13. Ectoloph of cheek teeth sinuous. 14. Parastyle buttress suppressed. 14. Parastyle buttress prominent. 15. Well-developed crochet and crista, united in 15. Crochet present but crista generally absent. worn tooth to enclose a medifossette. 19421 NOTES ON RHINOCEROS SONDAICUS. 2 3

Fig. 1. Comparison of the skull and jaw of (A) Rhinoceros unicornis, (B) Rhinoceros sondaicus and (C) Gaindatherium browni. Lateral views of right side showing, from C to B to A: (2) increase of nasal horn boss, (3) increase in height of ascending ramus, (4) increase in height and forward inclination of occiput, (5) increase in depth of "saddle" in cranial profile. Compare the items with these same numbers on page 2. Figures one-eleventh natural size A and B from Osborn, 1898; C from Colbert 1934.

I1 on "The Extinct Rhinoceroses" (1898) dis- works and original observations of the cussed briefly certain differences to be seen American Museum specimens of Rhin- between the living species of Rhinoceros, oceros, the accompanying comparison has and in addition presented a very useful been drawn up (see p. 2). comparative figure in which lateral views This comparison of the characters in the of the skulls and mandibles of all the living two modern species of Rhinoceros is inter- rhinoceroses were shown to scale. UJpon esting in that it shows what might be called the basis of these previously published a "harmonic" specialization of Rhinoceros 4 4AMERICAN MUSEUM NOVITATES [No. 1207 unicornis over Rhinoceros sondaicus. The 5. Shallower saddle in cranial profile. one species is advanced beyond the 6. Zygomatic arch more angular at posterior other termination. not by virtue of a few isolated characters 7. Posterior margin of palate with small but in all features throughout the structure median projection. of the skull, jaws and dentition; every 11. Premaxillaries narrow. (Two incisor teeth, character listed above for the distinction of a primitive character.) 12. Teeth more brachyodont than in R. sondai- the two forms shows an advance in the cus. Indian rhinoceros over its expression in the 13. Ectoloph of cheek teeth sinuous. Javan form. It is not possible on the basis 14. Parastyle buttress prominent. of the material available to compare skele- 15. Neither crista nor crochet present. tons of the two species, but it would seem It might be mentioned that in these likely that much the same picture would characters Gaindatherium in turn is more hold in a lesser degree in the post-cranial or less intermediate between Rhinoceros region. Certainly the external characters sondaicus and certain mid-Tertiary rhino- in the two species conform to this concep- ceroses such as Caenopus or Subhyracodon. tion of a general, all-round specialization The reason that Gaindatherium is thought of Rhinoceros unicornis over Rhinoceros to be directly ancestral to Rhinoceros is its sondaicus. possession of certain characters mentioned The validity and the direction of the in an earlier paragraph that distinguish the specializations in Rhinoceros, shown in two latter genus, specifically the presence of a successive stages by the modern species, single nasal horn and the consequent de- are corroborated by a study of the fossil ma- velopment of a saddle-shaped cranial pro- terial, particularly the evidence offered file, the presence of large, well-developed by the extinct genus Gaindatherium from incisors and the closure of the external the late Tertiary.phases of the Siwalik series auditory meatus inferiorly by the fusion of of India, described a few years ago by the the post-tympanic and post-glenoid proc- present author (Colbert, 1934). When it esses. was described, Gaindatherium was sug- At this place it might be well to consider gested as an ancestor to Rhinoceros, the briefly two Pleistocene species of Rhino- comparison being between the fossil form ceros, Rhinoceros sinensis and Rhinoceros and the Indian animal. With a skull of sivalensis, both known from considerable Rhinoceros sondaicus at hand it can now be suites of materials from China and India, seen how nicely the Javan rhinoceros fits as respectively. a form intermediate in position between Rhinoceros sinensis is a large species, Gaindatherium and Rhinoceros unicornis. almost as big as the modern Indian rhi- Every character that sets Rhinoceros noceros, but interesting in that it shows a sondaicus off from Rhinoceros unicornis combination of the characters that dis- is expressed with greater emphasis in tinguish R. sondaicus and R. unicornis. Gaindatherium. This is conveniently shown Thus, in spite of its large size, Rhinoceros when Gaindatherium is compared point by sinensis seemingly had a rather small horn point with Rhinoceros sondaicus, using the carried on a pointed horn boss, as in the numbers that were listed in the comparison, Javan rhinoceros. Since the back of the presented above, of the two living species. skull is not well preserved in this species, nothing can be said about the diagnostic Gaindatherium browni COMPARED WITH characters of the cranial and basicranial Rhinoceros sondaicus regions. However, the cheek teeth are 1. Gaindatherium is smaller and lighter than Rhinoceros sondaicus. interesting in that they are hypsodont, 2. Less expansion of the nasals; horn boss more so than the teeth of Rhinoceros pointed. sondaicus, less so than those of Rhinoceros 3. Ascending ramus lower. unicornis, they have the parastyle buttress 4. Skull generally lower. (The occiput of Gaindatherium is vertical, a primitive as in R. sondaicus but not so prominent, the character as compared with the for- ectoloph is less sinuous than in R. sondaicus wardly inclined occiput of Rhinoceros.) ^ but not so flattened as in R. unicornis, B

Fig. 2. Comparison of the glenoid and posterior palatal regions of the skull of (A) Rhinoceros unicornis, (B) Rhinoceros sondaicus, and (C) Gaindatherium browni. Palatal views of right side showing, from C to B to A:. (6) increase in angle between glenoid and zygomatic arch, (7) decrease in median projection on posterior border of palate,- (13) straightening of molar ectoloph, (14) reduc. tion of parastyle buttress, (15) development of crochet and crista. Compare the items with these same numbers on page 2. Figures one-third natural size. A and B from Osborn, 1898; C from Colbert, 1934. 5 6 AMERICAN MUSEUM NOVITATES [No. 1207 while the crochet and crista, although well velopment where they join to enclose a developed (the latter often being redup- medifossette. licated), do not join to enclose a medi- From these several comparisons certain fossette as in R. unicornis. Thus it seems facts emerge. Rhinoceros sondaicus may be evident that Rhinoceros sinensis is a form distinguished from Rhinoceros unicornis by of rather intermediate position between the a large series of characters in the skull, the two living species. mandible and the dentition. In all of these Rhinoceros sivalensis, on the other hand, characters the Javan rhinoceros is more would seem to be rather close to Rhinoceros primitive than the Indian form, so the unicornis. The extinct form is a large development of the one from the other species, showing in the structure of the may be thought of as a "harmonic" growth skull (the large horn boss, the deep saddle affecting virtually all portions of the cranial in the cranial profile, the forwardly in- and dental anatomy and proceeding in a clined occiput, the depth of the skull and single direction. Indeed, this development the like) a close approach to the modern in the characters from Rhinoceros sondaicus Indian form. The cheek teeth in this to Rhinoceros unicornis is so marked that certain Pleistocene species of the genus occupy positions intermediate between the twb modern forms, Rhinoceros sinensis being a truly intermediate type, Rhinoceros sivalensis being somewhat closer to the Indian species. Consequently, it would appear that Rhinoceros sondaicus, though recent in age, is truly a persisting primitive form and anatomically may be regarded as at about a Lower Pleistocene or perhaps an Upper Pliocene stage of development. It ap- proaches the Lower Pliocene genus Gainda- therium, being intermediate between this latter form and the more advanced Pleisto- cene and Recent species of Rhinoceros mentioned above. Therefore, Rhinoceros sondaicus upon the basis of present evidence is to be regarded as the most primi- Fig. 3. Phylogenetic development of the tive member of the genus Rhinoceros, em- genus Rhinoceros. bodying in its structure the features which by further complications and developments species are characterized by a rather flat became diagnostic for the Pleistocene ectoloph but by a certain retention of the species, Rhinoceros sinensis and Rhino- parastyle buttress, while the crochet and ceros sivalensis, and for the Recent Rhino- crista have not reached that stage in de- ceros unicornis. It is a true living fossil.

LITERATURE CITED COLBERT, EDWIN H. MATTHEW, W. D. 1934. A new rhinoceros from the Siwalik 1931. Critical observations on the phylogeny beds of India. Amer. Mus. Novitates, of the rhinoceroses. Univ. California No. 749, pp. 1-13. Publ. Bull. Dept. Geol. Sci., XX, No. 1, pp. 1-9. FLOWER, 'W. H. OSBORN, HENRY FAIRFIELD 1876. On some cranial and dental characters 1898. The extinct rhinoceroses. Mem. Amer. of the existing species of rhinoceroses. Mus. Nat. Hist., I, Pt. 3, pp. 96, 97, -- Proc. Zool. Soc. London, pp. 443-457. 116-119.