84 PRIMAkt' FACTORS OF ORGANIC EVOLUTION. 



They present the advantage, that, since they appeared 

 last of the animal kingdom in time, we can obtain a 

 clearer view of their beginnings than in the case of the 

 other great branches. 



Before going into the subject I wish to call atten- 

 tion to a prevalent source of error in the construction 

 of phylogenies. This is the confusion of ideas general 

 among naturalists who are not at the same time com- 

 petent systematists, as to the subordination of charac- 

 ters. All correct phylogenetic inference depends on a 

 correct appreciation of the value of characters. Fail- 

 ing this, error and confusion result. If, for instance, 

 it is alleged that such a genus is ancestral to another 

 genus, it is often forgotten that the descent of generic 

 character, and not specific character, is meant. The 

 usual type of critic attempts to contradict such hy- 

 pothesis by showing some incongruity in specific char- 

 acters, a matter which is quite irrelevant to the issue. 

 Thus Madame Pavlov finds that Hippotherium mediier- 

 raneum is not the ancestor of Equus caballus, and comes 

 promptly to the conclusion that the genus Hippothe- 

 rium is not ancestral to the genus Equus. This is a 

 -non sequitur, for there are perhaps twenty species of 

 Hippotherium, some of which are almost certain to 

 have been ancestral to species of American Equus. In 

 like manner, if it is alleged that the condylarthrous 

 Mammalia are ancestral to the Diplarthra, if it should 

 happen that no known genus of the former fits exactly 

 the position of ancestor to any genus of the latter, in 

 our present state of knowledge, the contention is not 

 thereby vitiated, and it is implied that such genus will 

 certainly be found. If it is also alleged that Condy- 

 larthra have been the ancestors of the anthropoid line, 

 if some of the known genera of the former turn out to 



