118 THE AMERICAN NATURALIST [Vol. XLII 



the process of splitting what was assumed by the older 

 zoologists to be a valid species into two or more new ones 

 has also considerably increased the number of described 

 forms. This practise, though sometimes questionable, is 

 at least illuminating, for it raises at once the fundamental 

 question of what constitutes a species. In the days of 

 Linnaeus when special creation was more generally ad- 

 hered to than now, it was comparatively easy to meet this 

 question by the statement that a species is the aggregate 

 of individuals represented by the originally created pair 

 or stock and their descendants ; but, with the acceptance 

 of the evolutionary idea, this reply no longer sufficed. 

 From the evolutionary standpoint every species has had 

 a history and this history has been clearly one of change 

 whereby the aggregate of more or less similar individuals 

 at one time representing the species gave rise to the self- 

 perpetuating stock whose more remote members evolved 

 in one or more directions new features, so that the species 

 either as a whole assumed new characteristics or split into 

 two or more subordinate groups, each having its own 

 special features and being destined eventually to become 

 as well circumscribed from its next of kin as the original 

 stock was at the outset. With this process in mind it is 

 fair to expect that nature would be found to embrace 

 many aggregates of individuals which would represent 

 species at all steps of differentiation, and whether the 

 individuals of a given aggregate had come to differ suffi- 

 ciently among themselves to constitute a new species or 

 not would depend entirely on the judgment of the natural- 

 ist who described them. It is thus clear that we can not 

 expect any fundamental characteristic by which a new 

 species can be definitely determined, for it is obvious that 

 the transformation of species is a more or less continuous 

 process in which the degree of separation whereby the 

 new species will be established is somewhat arbitrarily 

 determined by each describer. Hence the idea of species 

 rests upon an artificial basis, and, if the describer 's meas- 

 ure of specific difference diminishes with the progress of 



