THE COEFFICIENT OF MUTATION IN OENOTHERA BIENNIS L. 3 



dorsiventrality and of differentiation, and this without any refer- 

 ence to advantages or disadvantages to be derived from them in 

 their functions. Numerous points of similar significance in the 

 structure of the vegetative and reproductive organs are pointed 

 out by the author. Moreover, the genera Tristicha and Podo- 

 stemon, which are widely distributed, are comparatively little 

 modified from the earlier types of the orders, while the highly 

 specialized forms are at the same time the rarest, exactly as in 

 the case of the endemics of Ceylon. 1 ) 



In the group of the evening primroses the same principles 

 prevail. Their struggle for existence is limited by the difficulties 

 which they have in producing roots. Cuttings almost never suc- 

 ceed in rooting, with the exception of the lateral rosettes at the 

 base of the stem. Artificial transplanting becomes difficult as 

 soon as the main root increases in size. In the field only a small 

 percentage of the seeds germinate and thrive, and this only under 

 special conditions. They want a stirred up soil and do not like 

 to grow between other plants. These characters are common to 

 all the forms which I have had an opportunity of studying in their 

 native habitats. On the other hand, the numerous small specific 

 differentiations, such as the form of the leaves, the branching of 

 the stem, or the structure of the flowers and fruits, do not show 

 the least relation to their environments. Even the preference for 

 an annual or a biennial behavior, which might seem to be a direct 

 adaptation, does not exhibit any reference to the actual life condi- 

 tions. The conception of natural selection and of the accumu- 

 lation of small variations on account of their utility cannot explain 

 the specific and generic differences in this group. 



Therefore it seems unavoidable to assume that specific differ- 

 entiation in the genus Oenothera has been produced and is still 

 being produced by small steps, each of which evolved a character 

 at once to its full development, without any reference to the struggle 

 for life. That, besides this process, from time to time new com- 

 binations of characters by means of crosses may have given rise 

 to constant hybrid strains, which we have as yet no means of 

 distinguishing from pure species, cannot of course be doubted. 



Now, Oenothera Lamarckiana, 0. biennis, and some allied forms 

 are seen to be still in a condition of making, from time to time, 



i) Willis, J. C, On the lack of adaptation in the Tristichaceae and 

 Podostemaceae. Proc. Roy. Soc. 8:532 — 550, 19 14. 



I* 



