PROMERYCOCHCERUS 105 



diversified and that at least some of the species were undoubtedly directly ancestral to certain species 

 of Promerycochoerus. Loomis has indicated a very considerable amount of sexual dimorphism and 

 many age changes. My belief is that, while there is much difference due to these factors, the great 

 amount of variation lies deeper, in the ancestry of the group. Perhaps we should subdivide the 

 genus, but in our present state of knowledge I consider this to be inadvisable. When we know 

 more skeletons of the species, we can make an intelligent classification and then combine species, 

 possibly separate the sexual characters, and lay down a definite basis for the arrangement of these 

 numerous forms, the great majority of which are known only from skulls. I cannot be sure which 

 are males and which females, other than to consider that the female has a somewhat lighter, narrower 

 skull and longer jaws. It is on this basis, which may not be valid, that I have placed M or F 

 following the species name in the list given below. 



Any classification at present must be more or less arbitrary, and I am not dogmatic about the 

 rearrangement below. This is based on comparisons of the type and other material, but it must be 

 kept in mind that of the more than twenty species described only four have a complete or nearly 

 complete skeleton, and these four intergrade from heavy, long-bodied animals to those with shorter 

 bodies and lighter, somewhat longer limbs, thus leaving us with no clear-cut distinctions that might 

 be used as a basis for subdivision of the genus. The body lengths vary to an extent of about 

 400 mm. but the height only to 1 1 5 mm. 



Undoubtedly there is synonymy in this group. I have indicated where I believe this may lie, 

 but only future discoveries can enable us to finish this chapter. 



The two California species, P. erythroceps, from the Tecuya beds, and P. hesperus, from the 

 upper Sespe, consist only of the anterior part of the muzzle and of but a fragmentary right ramus 

 respectively. That is, no parts common to both are available for direct comparison. The former 

 species is supposedly geologically older than the latter. In several respects both species bear a 

 resemblance to P. leidyi but also partake of a few of the characters found in P. chelydra and P. hol- 

 landi hatcheri. Due to the paucity of material, we shall consider these two species valid for the 

 present. When we know more about them, I am inclined to believe that these two forms will prove 

 to be very much alike, representing a distinct fades in that area, and possibly the older giving rise to 

 the younger. 



The John Day species show a large variety of skull characters, but no complete or nearly com- 

 plete skeleton from this area has been described. P. superbus, the genotype, as well as the other 

 species shows very slight differences in individual tooth proportions, although the lengths of the indi- 

 vidual dental series and their indices vary much. I believe that P. chelydra latidens is probably the 

 male of P. chelydra, while P. macrostegus, P. m. infatus, and P. marshi represent an allied group, 

 having a general similarity but showing as well many variations. P. leidyi is not far from P. super- 

 bus. P. curvidens, of the upper Oligocene, may well have been the ancestor of P. leidyi, as their 

 general proportions are much alike. The former is much smaller and may represent a dwarf variety. 

 P. lulli and P. microcephalus, also upper Oligocene, differ from each other and from the other 

 John Day species. 



Any arrangement of the Great Plains forms is also problematical. I should place the female of 

 P. thomsoni as a slenderer form of P. carrikeri but closely allied. P. thomsoni M and P. vantassel- 

 ensis are very similarly proportioned, and the former in the Rosebud may well have given rise to 

 the latter in the Harrison. With these I should put P. v. pygmceus as a dwarf variety. 



There is a peculiar character shown in the tooth dimensions of P. gregoryi F, P. hollandi, 

 P. h. hatcheri, and P. h. minor in that the inferior molar-premolar indices are lower than the 

 superior. In all of the other species of the genus the reverse is true. There are other similarities 

 as well. P. gregoryi F, of the Rosebud, could well have been ancestral to P. hollandi, and P. h. 

 hatcheri may be the female form of the latter, while P. h. minor I consider to be the dwarf variety of 

 the P. hollandi stock. 



P. montanus (type) lacks the muzzle and is otherwise much damaged, and its exact horizon is 

 not definitely known. To me, however, P. montanus and P. m. grandis seem very much alike. The 

 former is perhaps the female form of P. m. grandis, or, if P. montanus is really of upper Miocene age, 



