'600 UNGULATA. 



to the merest vestige of its metacarpal, the pes is very similar to that of 

 the preceding family. The changes here, as compared with the previous 

 family, are the presence of ridges connecting the tooth-cusps, the reduction 

 of digit No. 5 of the manus. 



We now come to the Equidae, where we find Protohippus with hypsodont 

 molars, about half the length of those of Equus, and valleys filled in with 

 cement, the anterior inner tubercle of the tipper molars is connected with 

 the intermediate tubercle, orbit closed, manus and pes tridactyle, the 

 outer digits not reaching the ground, no trace of any other digits, ulna 

 slender distally and fused with radius. Lastly we have Pliohippus with 

 rather longer grinders, very similar to Equus in pattern, without lateral 

 digits, the metacarpals only being represented ; ulna and fibula still more 

 approximating to the condition found in Equus. 



So far as the characters mentioned are concerned, we have here a 

 very remarkable series of forms which at first sight appear to constitute a 

 linear series with no cross-connections. Whether, however, they really do 

 this is a difficult point to decide. There are flaws in the chain of evidence, 

 wh'ch require careful and detailed consideration. For instance, the genus 

 Equus ape ears in the Upper Siwalik beds, which have been ascribed to 

 the Miocene age. It has, however, been maintained that these beds are 

 really Lower Pliocene or even Upper Pliocene. It is clear that the decision 

 of this question is of the utmost importance. If Equus really existed 

 in the Upper Miocene, it was antecedent to some of its supposed ancestors. 

 Asain in the series of equine forms, Mesohippus, Miohippus, Des- 

 mathippus, Protohippus, which are generally regarded as coming into 

 the direct line of equine descent, Scott * points out that each genus is, in 

 some respect or other, less modernised than its predecessor. In other 

 words it would appear that in this succession of North American forms 

 the earlier genera show, in some points, closer resemblances to the modern 

 Equus than to their immedate successors. It is possible that these 

 difficulties and others of the same kind will be overcome with the growth 

 of knowledge, but it is necessary to take note of them, for in the search 

 after truth nothing is gained by ignoring such apparent discrepancies 

 between theory and fact. 



Fam. 5. Rhinocerotidae.f Large unwieldy pachyderms, usually with 

 one or two epidermal horns on the strongly arched nasal and frontal 



bones ; i ~~ c \~^ P 4 ^| m f , complete only in the oldest types ; 

 incisors and canines frequently absent ; premolars like the molars but 

 simpler in the older forms ; last lower molar without third lobe ; u. grinders, 

 with thick outer wall, nearly flat with obliquely transverse laminae, lower 

 grinders with two semilunar ridges joining to form the outer wall. The 

 orbit is widely open behind and the frontal is without a postorbital process ; 

 the nasals are extensive ; the postglenoid processes very large and may or 

 may not unite with the post-tympanic process of the squamosal to form a 

 false external auditory meatus ; the tympanic is annular. Ulna and 

 radius, tibia and fibula complete. Manus with 3 or 4, pes with 3 digits ; 

 digit No. 3 is larger than the others and symmetrical in itself, digit No. 1 

 is not present, and digit No. 5 when present in the manus is smaller than 



* Trans. American Philosophical Society (N.S.) 18, 1896, pp. 119, 120. 



f Lydekker, Notes on Rhinoceroses ancient and modern, Field, 79, 

 p. 903,' and 80, p. 38, 1892. Pavlow, Les Rhinoceridae de la Russie, etc., 

 Bull. Nat. Moscow, 1892, p. 147. 



