1 8 HiCKLiNG, Variation of Planorbis miiltifonnis, Bro7Z7t. 



preponderance of " high " forms is the result of a selective 

 death-rate during growth. This involves the assumption 

 of an enormous range of purely simple variation, for which 

 no explanation can be offered, while it takes no account 

 of the geological history of the species. The hypothesis 

 here put forward might be tested by an accurate statis- 

 tical investigation to determine the exact form of the 

 variation curves. Indeed, this suite of shells would seem 

 to offer an exceptionally favourable field for the statistical 

 investigation of a variety of problems connected with 

 variation and inheritance. 



A few other points seem to require brief references. 

 Those who are familiar with Hyatt's paper will have 

 recognised that the forms here described represent only 

 the third and fourth of the four genetic series which he 

 has constructed — possibly the fourth only, though I think 

 the third should be united with it. No representatives of 

 his first and second series are present in my material, and 

 I am therefore not in a position to discuss the relations of 

 these. The earlier members of the third and fourth are 

 likewise unrepresented by adult shells. His fourth series 

 begins with Planorbis Steinheitnensis, a form closely in 

 agreement with the theoretical ancestor which ontogeny 

 had led us to anticipate — the " non-carinate discoid " type. 

 It has been shown that among the shells here considered 

 this type no longer exists, though it is practically repre- 

 sented by the young stages of some individuals. Never- 

 theless the material does actually contain a few shells 

 which are either P. Steinheiiiiensis or a very slightly 

 modified form derived from it {PL II., Fig. 8) ; but these 

 do not belong to the series. There are absolutely no 

 intermediate forms to connect them with the others,. 

 except among the extremely young shells. Here, then, we 

 seem to have a type which has now definitely separated 



