Cope.] * t [May 16, 



part or piece of the organism.* If now we fail to notice these points or 

 steps, we must abolish all genera. If we define some and fail to define 

 others, our practice ceases to -have the uniformity of a law, and we aban- 

 don the basis of scientific order. f One point, however, must be insisted 

 on. In order that a character be usable for any purpose of definition, it 

 must define. That is, it must belong to all the individuals referred to the 

 species, genus, etc., defined, and must not be present in some individuals 

 and wanting in others of those supposed to be defined by it. This being 

 the case, adult animals only can be used for definition, as characters, es- 

 pecially generic, are added from time to time up to maturity. Sometimes 

 Only one 'sex can be considered, since the adult characters are in certain 

 cases never reached by one sex or the other. This is often the case with 

 insects. Moreover, some latitude for exceptional variations must be al- 

 lowed. Thus, the exceptional absence of the last molar in a dog does not 

 invalidate the definition of the genus Canis, M. f . 



Of course, if all specimens of animals could be found, the definitions 

 would all, or nearly all, be invalidated. But it is safe to assume that all 

 the intermediate forms will not be found, so that the definitions of species 

 will represent the state of our knowledge, and the results of the operations 

 of nature in the preservation of individuals. 



The case is somewhat different with regard to generic characters. As 

 these involve the addition or subtraction of some part, having definite 

 dimensions, it is quite possible to say when the latter is present or absent. 

 Characters of this kind present the appearance of abruptness of transition, 

 to which I have referred in my paper "On the Origin of Genera," and 

 which gave rise to the formulation, by Professor Hyatt and myself, of the 

 "laws of acceleration and retardation." When such change prevails 

 throughout all the individuals of one or more species, a new genus has its 

 origin. As a matter of fact, the creation of generic modifications has been 

 exhibited, in the history of lite, by many individuals nearly contempora- 

 neously. As the change involves but one character, it offers a better oppor- 

 tunity for the formulization of the laws of evolution, than in the case of 

 specific characters, which are more numerous. 



The three genera of Elephantidse, of which mention has been made 

 above, will then be defined as follows : 

 Mastodon Cuv. Superior incisors without enamel band ; inferior incisors 



wanting. Type M. americanus. 

 Dibelodon Cope. Superior incisors with enamel band ; inferior incisors 



wanting. Type B. shepardi. 

 Tetrabelodon Cope. Superior incisors with enamel band ; inferior incisors 



present in the male at least. Type T. angustidens. 



To the genus Mastodon must be referred the following species. For 



* See "Origin of Genera," Proc. Acad. Philada., 1869, where this point is dis- 

 cussed, 

 t American Naturalist, 1884, July, p. 



