NO. 8 UPPER EOCENE ARTIODACTYLA — GAZIN I3 



as the directions taken in dental specialization, though similar as to 

 adaptation, are entirely different in detail. Both are large bunodont 

 forms, but in Archaeotherium the snout is very elongate, all the pre- 

 molars are retained, and the secondary cuspules, lost or reduced in 

 Achaenodon, are as much emphasized as the primary cusps. If 

 Archaeotherium is related to the Helohyinae, it is through an earlier 

 form, possibly Lophiohyus, possessing a dental structure and formula 

 that had not gone too far in the direction of Achaenodon. I sug- 

 gest Lophiohyus only because this form had already developed an 

 elongate snout. Its dentition is very much like that of Helohyus. 



The Tayassuidae is shown in the chart, not because of any particular 

 conclusions as to origin in the dichobunids, but because it might well 

 have been derived from North American forms. However, in this 

 instance, the European cebochoerids must not be overlooked as po- 

 tential ancestors. Its origin may lie somewhere in the middle Eocene 

 homacodonts rather than in the helohyids as suggested by Pearson, in- 

 asmuch as the posterointernal cusp of the upper molars in Perchoerus, 

 unlike Helohyus, would appear to be the hypocone, and as the 

 metaconule is much reduced — that is, unless the small cusp in the 

 position of a metaconule in Perchoerus was newly acquired and not 

 the original metaconule. 



Directing our attention to the selenodont families, we find the 

 agriochoerids and oromerycids, though with highly crescentic appear- 

 ing molars, somewhat intermediate between the dichobunids on the 

 one hand and the leptomerycids and poebrotherine camelids on the 

 other, in the perhaps retarded or differently developed protocone above 

 and hypoconid below. For the agriochoerids I have shown a rather 

 simplified version of their relationship in the general chart, but their 

 lineage within upper Eocene time may have been somewhat more 

 complex in detail, as suggested in chart 2 (p. 67). I do not believe 

 that more than two genera are represented, but within Protoreodon 

 there appears to have been at least two distinct sequences of species 

 or mutants that can be carried from middle through upper Uinta 

 levels. The larger of these was evidently derived from or was close to 

 P. parvus and carried through Uinta C as P. pumilus, culminating in 

 P. pumilus annectens. This may well be the line that gave rise to 

 Agriochoerus. A sequence of smaller forms can be recognized in 

 P. minor of lower Uinta C, culminating in P. petersoni in upper 

 Uinta C. If Merycoidodon was derived from any of the known forms 

 of Protoreodon, the line represented by P. petersoni is surely the most 

 likely source. Nevertheless, this is an agriochoerid that cannot be 



