NO, 8 UPPER EOCENE ARTIODACTYLA — GAZIN II 



for which there is no supporting record, this suggestion may seem 

 highly speculative but the resemblance must be more than coincidence. 



In conflict with the foregoing hypothesis attention should be called 

 to the reduction of the lateral toes of the hind foot to slender vestiges 

 in certain material of Diacodexis, a condition that led Matthew to 

 believe that Diacodexis could not be ancestral to any of the later 

 Artiodactyla. However, the extent to which this condition may be 

 variable in the Diacodexis complex, in view of the much better de- 

 veloped lateral digits of the fore foot, is not known, and Diacodexis 

 certainly shows striking variability in its dentition. Moreover, the 

 hind-foot structure of Leptochoerus seems obscure since Scott, as 

 late as 1945, referred to Marsh's statement that in a Yale specimen, 

 "The hind foot resembles that of Homacodon, having four usable 

 digits, but the navicular and cuboid are co-ossified, an unexpected 

 feature." 



The second dichobunid subfamily group, for which the name 

 Homacodontinae is retained, was certainly not derived from Diaco- 

 dexis or the Diacodexinae. The Wasatchian form Hexacodus, which 

 has for its counterpart Protodichobune from the lower Eocene of 

 d'Epernay, France, would appear to have been more nearly in the 

 line of descent for later typical homacodonts but was evidently off to 

 one side and, though much less specialized in the dental peculiarities 

 characterizing Antiacodon, may well have given rise to that Bridgerian 

 genus. Nevertheless, it is worth noting that within the upper Wa- 

 satchian material of Hexacodus a certain variant in the dental pat- 

 tern is strikingly suggestive of Microsus. Microsiis and Homacodon 

 are rather alike although the former has teeth with cusps that appear 

 to be more acute. The teeth of Microsus are of more delicate and 

 perhaps more primitive appearance and potentially better suited to 

 the ancestral position with respect to the upper Eocene homacodonts 

 than is the large Homacodon vagans. Unlike Homacodon, the fourth 

 lower premolar of Microsus has a pronounced metaconid, but this 

 does not preclude a relationship suggested by the chart. 



By upper Eocene time the homacodonts, though still essentially 

 bunodont, had acquired rather marked styles on the outer sides of 

 the upper teeth, and the outer cusps of the lower molars had become 

 a little more crescentic. In one line, represented by Bunomeryx and 

 later Pentacemylus, the hypocone was early lost and a mesostyle early 

 gained in the upper molars. Mytonomeryx is apparently related but 

 was derived from the stock before the hypocone became reduced. It 

 further specialized in a lengthening of the snout. Hylomeryx likewise 

 retained the hypocone but did not achieve a mesostyle. As a remain- 



