No. 484] 



CLOSELY RELATED SPECIES 



231 



ters. In order to avoid complications I have used the word kind 

 to designate such different types, instead of the words species, 

 variety, etc., which have restricted technical senses. 



I have sought for closely related pairs of kinds so made up that 

 in each case no third kind stands between the members of the 

 pair in resemblance. Such pairs I may call immediately cognate 

 pairs, or for short, cognate pairs. A pair may consist of two 

 species, two varieties, two subspecies, a species and a subspecies, 

 a species and a variety, etc. It is assumed that such cognate 

 pairs represent recent forkings of the phylogenetic tree; and that 

 if we could collect all such cognate pairs in the vegetable kingdom 

 we should have a representation of all the youngest forkings. 

 Evidently their distribution would be very illuminating, for the 

 youngest branches are on the average the least disturbed geo- 

 graphically, and the distribution of the members of these pairs 

 would represent as accurately as we could ever discover it, the 

 position of things at the moiiient when forking takes place. That 

 is, we should liave a ucouiaphic cliart, more or less distorted it 

 is true, of the origin of kinds. If the members of the pairs are 

 universally, in the vegetable kingdom, separated from each other, 

 then — as already explained — Mutation is excluded as a true 

 cause of diversification of hereditary types in plants. For among 

 several forms of isolation to which Mutation may conceivably 

 give rise, and which are, therefore, not inconsistent with the mu- 

 tational assumption, the one form of isolation to which it could 

 never give rise is geographic isolation. 



I repeat that I have examined only the broad geographical 

 aspect of distribution and not at all the topographical, for which 

 exact data are wanting. Let the reader recall the two stages of 

 this general inquiry: my evidence belongs to the first of these. 

 I present the following facts as a contribution towards an answer 

 to the question. Is Mutation instantly excluded from a place among 

 the considerable powers in evolution, by the broad aspects of specific 

 distribution in plants? I have taken only one step. But this 

 may be of some little importance, especially in view of the asser- 

 tions concerning the distribution of plants which have been made, 

 and in view of the lack of even broadly geogi-aphical statistics. 



