353 



ON THE VARIATIONS OF THE EVENING PRIMROSE 



(CENOTHEBA BIENNIS L.). 



By G. A. Boulenger, F.R.S. 



Most naturalists have taken a great interest in the observa- 

 tions of De Vries on the variations of CEnothera, as observed by 

 him in Holland, which have been summed up in his classical 

 work M-utationstheo) ie, but few as yet seem to have gone to the 

 trouble of forming an opinion of their own on the degree of dis- 

 tinctness of the various "mutants" which are claimed to repre- 

 sent new "elementary species." The question whether or not 

 new species appear at the present day is one of the greatest 

 importance to the philosophical biologist, and if it can be shown 

 that De Vries is mistaken in his conception of " species," it 

 remains as it stood before. 



Many are the localities where the Evening Primrose has 

 established itself in this country, and it would be of very great 

 importance to test at different spots the results arrived at by 

 the celebrated Dutch naturalist, an easy and interesting piece of 

 work for our local botanists. 



I have had an opportunity of looking into the matter at two 

 different places, and I have found the results so different from 

 those obtained by De Vries, that I think it desirable to place my 

 observations on record, as an incentive to others to pursue the 

 investigation. 



To begin with, I have not been able to confirm the distinction 

 established by De Vries between CE. Lamarckiana and CE. biennis, 

 which I find to be connected by every possible transition, the 

 characters, based on the flower, on which he distinguishes the 

 species being of the order of continuous variation. Further, in 

 observing the behaviour of (E. Lamarckiana in a sandy locality 

 on the French seaside, I do not find a species producing here and 

 there a few mutants, but an assemblage of specimens showing 

 endless variation in every direction. If species were to be esta- 

 blished on the basis adopted by De Vries, there would be an end 

 of this aspect of systematic work. I feel fully convinced that the 

 experienced systematists who have preceded De Vries were right in 

 regarding CE. biennis as a highly variable species, the forms of 

 which could hardly be defined as well-marked varieties. When 

 De Vries argues that his new species are as satisfactorily esta- 

 blished as those of many authors, he is right only in so far as 

 the species to which he alludes were proposed by uncritical 

 botanists upon herbarium specimens without any knowledge of 

 the variations displayed by these plants if studied in the open on 

 a large material. 



It has been thought before that CE. Lamarckiana is only a 



garden variety of CE. biennis ; also that it may be a hybrid, al- 

 though, as observed by MacDougal, it has not been suggested with 

 what species CE. biennis would have hybridized to produce (E. 



Journal of Botany. — Vol. 45. [October, 1907.] 2 d 



