20G PROCEEDINGS OF THE ACADEMY OE [April, 



(as it probably is in most cases of sexual dimorphism in the lower 

 animals). If both sexes had a parallel ontogenetic development, 

 we should not hesitate to classify the individuals at the terminal 

 stage of their reproductive activity, whether the development had 

 been progressive throughout or in part regressive. AVhen there has 

 been a divergence in the ontogeny of the two sexes, the males and 

 females being complemental individuals, the species contains both 

 jirogressive and regressive individuals, it is to be classified at the 

 time of repi'oductive activity of the two sexes, and is to be ascribed 

 a rank intermediate between that of the conditions of the two sexes. 



Unless this principle of considering all the kinds of individuals in 

 a species be strictly enforced, great confusion woidd ensue in the 

 ranking of a species, depending as to whether the classifier con- 

 sidered the males alone or the females alone, or asexual individuals 

 alone, or all together. It is necessary that all the individuals of a 

 species be knoAAni befox-e such a species can be thoroughly character- 

 ized, and for only a comparatively few species of certain animal 

 groups have such broad definitions been given. 



It will of course be understood that the writer does not imply that 

 species would be perceptibly delimitable, were it not for the extinc- 

 tion of or our lack of knowledge concerning intermediate species. 

 ' * Species " is a mental concept for a primary group of individuals 

 of the closest genetic connection. If in any phyletic series we could 

 know to-day all intermediate stages, so that the Avhole would be a 

 continuous line of development, and assuming that such a develop- 

 ment were perfectly gradual, nevertheless we should have to project 

 the concepts of species into that whole, arbitrary though such con- 

 cepts would be, in order to secure certain fixed points for observation 

 and interpretation. Thus we study a section through the body of 

 an animal if we cannot understand this animal in its entirety, and 

 by supplementing our knowledge with sections from other regions of 

 the body gradually reconstruct the whole. A species is a mental 

 section of a line of evolution, and is necessary to enable us to inter- 

 pret the whole. That is Avhat is here understood by the determina- 

 tion of a species, AVhere, through extinction, we find disconnected 

 species, we have sections made for us by Nature. 



V, The Kinds of Homolocjies. 

 Before considering the methods of treatment of structural char- 

 acters, and comparing the anatomical and em bryological methods of 



