﻿BOTANICAL GAZETTE 



described by Brixton, has been found only from Vermont, eastern 

 Massachusetts, and New York to the mountains of North Carolina. 

 The relative distribution of these species is similar to that of the 

 species of Clintonia previously considered, although in this case R. 

 allegheniensis occurs wholly within the range of the other species. 

 The specific differences may be tabulated as follows: 



R. abortivus Linn. R. allegheniensis Britton 



S tern leaves divided into oblong Stem leaves divided into linear 



or linear, somewhat cuneate acute segments 



lobes Stem glaucous 



Petals pale yellow, shorter than Petals pale yellow, minute 



the small reflexed calyx 



Styles very short, curved Styles subulate, hooked, 

 nearly half as long as the 



The main distinguishing feature of these two species is the 

 conspicuous recurved beaks of the achenes in R. allegheniensis. 

 The other differences are very inconspicuous and in themselves 

 scarcely noticeable. It seems reasonable to suppose that this 

 species has arisen from R. abortivus through a single positive 

 mutation. The idea that these conspicuous beaks might have 

 been gradually developed through natural selection might have 

 been readily accepted at the end of the last century, but has since 

 lost its plausibility. All the facts, both of characters and distribu- 

 tion, are more reasonably explained on the mutation hypothesis. 

 An increased length of beak is, so far as I know, of no considerable 

 use to the plant, although it is possible that the large hook might 

 aid the seeds in transportation by attachment to animals. In 

 distribution, however, the plant, while locally abundant, is restricted 

 in area, and R. abortivus surrounds it on all sides except where they 

 both reach the Atlantic coast. This points, not to its having an 

 advantage over R. abortivus in the struggle for existence, but more 

 probably to its having originated from that species relatively 

 recently through a mutation, and having since propagated itself 

 and spread with no conspicuous advantage or disadvantage in com- 

 petition with the parent form. 



The wide northerly distribution of R. abortivus makes it appear 

 probable that it is the older species and has given rise to R. alle- 



