June 5, 1903.] 



SCIENCE. 



913 



shown by skulls in the American Museum, are 

 intermediate forms and quite closely allied. 

 The height of sagittal crest, assigned as one of 

 the distinctive generic features of Daplicenus, 

 is a highly variable character in most car- 

 nivora, dependent on sex, age and individual 

 robustness. A series of opossum skulls will 

 well illustrate analogous variations, as recently 

 described by Allen. Serrations are to be 

 found on the unworn canines of all daphsenoid 

 dogs that I have examined, but disappear 

 very quickly with wear. Canines of old ani- 

 mals are smooth and more rounded in section 

 from wear.) 



As the names indicate, Mr. Hatcher be- 

 lieves that Proieinnocyon is ancestral to 

 Temnocyon of the John Day formation, and 

 Proaniphicyon to Amphicyon of the Loup 

 Fork, while Daphwnus left no descendant. 

 Scott, Eyerman, and Wortman and Matthew 

 had, on the contrary, derived Temnocyon from 

 Daplicenus, and all previous authors have re- 

 garded Amphicyon as a distinctively Euro- 

 pean type which found its way to America 

 only in the later Miocene. 



Mr. Hatcher does not recognize Mesocyon 

 Scott (^ Hypoiemnodon Eyerman, type 

 Temnocyon corypliceus Cope) as a valid genus, 

 and bases his comparison of Protemnocyon 

 with Temnocyon upon T. corypliceus, and not 

 upon the typical species (T. altigenis and T. 

 ferox). The authors above mentioned had 

 derived the typical Temnocyons from 

 Daplicenus but threw out Mesocyon corypliceus 

 from this line of descent. 



(Mr. Hatcher can hardly have seen Dr. 

 Eyerman's paper of May, 1896, for he could 

 not fail to observe that the characters as- 

 signed to separate Temnocyon and Mesocyon 

 are identical with those by which he separates 

 Daplicenus and Protemnocyon, only they are 

 even more marked and certain differences in 

 the teeth are superadded. In the White River 

 there are intermediate species between the 

 two extremes; in the John Day these have 

 not been found. If then Protemnocyon is a 

 good genus, Mesocyon must certainly be held. 

 If we can assume that the John Day forma- 

 tion is of later age than the White Eiver, it 



appears probable that Mesocyon and Temno- 

 cyon represent the further progress of the 

 difEerentiation between the large-skulled ro- 

 bust Daphwnus and the small-skulled, more 

 slender Protemnocyon. The extremes have 

 become more divergent and the intermediate 

 forms weeded out. The Daphcenus-Temno- 

 cyon line appears to lead into a type such as 

 Cyon, or the dholes, and evidences of an in- 

 termediate stage from the Loup Fork Mio- 

 cene were described by Matthew about a year 

 ago. The Protemnocyon-Mesocyon line leads 

 into much more typical dogs, but can not be 

 considered as a direct ancestor of any living 

 species which I have examined. 



Mr. Hatcher's derivation of Amphicyon 

 americanus from Proamphicyon is, I think, 

 hardly admissible. Amphicyon first occurs 

 in America in the upper Miocene Loup Fork, 

 but in Europe it is found in the oldest Oligo- 

 cene formations, as old as or older than the 

 White Eiver. The evidence is not at all such 

 as to warrant our affirming the actual con- 

 vergence of the Miocene Amphicyons of 

 Europe and America, the one derived from 

 one Oligocene stock, the other from a widely 

 different one. We might, perhaps, believe 

 that Proamphicyon and the European Oligo- 

 cene Amphicyons had a common Eocene an- 

 cestor; but as Proamphicyon is in fact very 

 much nearer to Daplicenus than to Amphicyon 

 it seems more reasonable to suppose that the 

 latter is, as Wortman believes, derived from a 

 distinct group of short-jawed dogs of the 

 Middle Eocene.) 



Mr. Hatcher makes at the close of his 

 memoir some good-natured criticisms of the 

 views expressed by Wortman and Matthew in 

 1899 as to the ancestry of certain Canidse.' 

 That such phylogenies are to a high degree 

 hypothetical, and seldom, if ever, more than 

 approximations to the truth, I am most ready 

 to admit — and have always regarded such a 

 saving clause as implied in any phylogenetic 

 remarks. But the new evidence brought for- 

 ward since then by Wortman and myself, and 

 now by Mr. Hatcher, serves to confirm in 

 most points the very lines of descent which 

 we suggested at that time. 



W. D. Matthew. 



