MODES OF MAMMALIAN EVOLUTION 657 



in a very large number of cases, but whether it is always valid 

 is very doubtful. In the Guinea Pig, as in all its family 

 (Caviidae), there are four toes in the front foot, three in the hind ; 

 but Professor Castle has lately succeeded in producing a race 

 with four toes in the hind foot. To call this a "monstrosity" 

 or ''abnormality " explains nothing ; the fact remains that the 

 four-toed race has been estabhshed and no reason can be 

 assigned why the same thing might not happen in nature. If 

 Dr. Matthew's view concerning the origin of the American 

 deer from fLeptomeryx (p. 409), should prove to be well founded, 

 another example of the same kind would be furnished. In 

 "fLeptomeryx of the Oligocene the upper canine was reduced 

 to minute, almost vestigial proportions, while in the ancestral 

 deer, ^Blastomeryx of the lower Miocene, it was a large, scim- 

 itar-like tusk. WTiile I am unable to acept this derivation of 

 the deer, it may be true nevertheless and, if so, will be a most 

 interesting example of the rehabilitation of a vestigial organ. 

 Decision must await the discovery of the intermediate forms. 

 Many such cases and instances of the addition of parts may be 

 so far undetected, but the phylogenetic series, as we have them 

 before us, point decidedly to the conclusion that such rehabilita- 

 tion or new addition is exceptional. 



III. So far as we are able to follow it by the aid of the fossils, 

 development among the mammals would appear to be a re- 

 markably direct and unswerving process. When any long- 

 lived phylum, made up of numerous well-preserved members, 

 is studied, the observer cannot fail to be impressed by the 

 straightforward course of the evolutionary process, as though 

 the animals were consciously making for a predetermined goal, 

 which, needless to say, they were not. A minute cusp makes its 

 appearance on a tooth, enlarges steadily in each succeeding 

 genus, and ultimately becomes a very important element in 

 the pattern ; and in this series of changes there is no oscillation 

 backward and forward. In the perissodactyls and a few other 

 groups, the premolars in each family gradually and steadily 

 2u 



