﻿i 7 8 



BOTAXICAL GAZETTE 



topographical conditions. In passing from east to west or from 

 north to south of the continent, the succession of species often 

 seems to form a graded series, though more or less disturbed and 

 modified by the incidence of mountain ranges, plains, etc., a fact 

 which is relied upon in many ways by systematic botanists in 

 their pursuits. 



This preliminary study of the relationships and distribution of 

 pairs of species is perhaps of more value for its suggestiveness than 

 for any direct contribution of facts; at least it is to be hoped that 

 such is the case, for the species I have chosen are all familiar forms in 

 the North American flora. Nevertheless, from the material I 

 have examined in this study it has been necessary to describe several 

 new and hitherto unrecognized varieties, and certain others will be 

 described in another connection. 



Notwithstanding the great amount of speculation, and more 

 recently of experimental work, on the factors of evolution, scarcely 

 any attempt has been made hitherto to show how one living 

 wild species has been derived from another particular species, or 

 from the common ancestor of both. No doubt systematists fre- 

 quently have such questions in mind when delimiting species, but, 

 if the methods of experimental evolution are sound, they should 

 enable us by now to begin the application of the ideas so gained to 

 the solution of simple examples taken from wild nature. Even 

 though the historical relationships of many species must remain 

 obscure, yet there exist cases in which the course of events is simple 

 and to some extent within our present powers of analysis. 



The pairs I have chosen have been taken at random. In a 

 subsequent study I may make a more methodical selection. In 

 some instances of species pairs the genus is bitypic; in others the 

 two species may stand apart from the others in the genus, either in 

 their structure or their distribution. Some of the cases of real 

 pairs, however, are not obviously pairs at all, and are only found 

 to be such by a study of their internal structure. On the other 

 hand, some species which form an apparent pair in a given region 

 are not very closely related to each other, and have only become 

 paired through the vicissitudes of altering distributions. Such 

 instances show that the mere taxonomic comparison of species, un- 



