dence, which are too involved to discuss here, tending to show that 

 the chromosomes are not all alike in their relation to heredity, hut 

 that they are, as we say, qualitatively different. Therefore a germ 

 cell lacking both members of a given type of chromosome (one de- 

 rived from each parent) would be lacking, in the ability to produce 

 the corresponding set of characters. You thus see that in these 

 occasional irregular distributions of chromosomes in the germ cells 

 in the Evening Primrose ye have a possible basis for the sudden 

 appearance of apparently new sets of characters which are, how- 

 ever, believed to be due to the loss and not to the addition of 

 anything. 



It should be borne in mind that this view of mutation as a 

 process of analysis is equally^ valid whether it be based upon the 

 chromosome distributions or considered entirely apart from these 

 phenomena. 



]ft seems probable that much species-formation may have taken 

 place in the manner I have suggested, by an analytical process in 

 which certain factors are lost from the germ plasm of the parent 

 species, thus giving rise to a series of different types. But this evi- 

 dently can not be the only evolutionary factor, for quite a different 

 set of forces is necessary to account for the origin of new organs 

 and for the steady progression in complexity which has taken place 

 so many times in the evolution of the plant and animal kingdoms. 



In conclusion I should like to outline to you a study I am mak- 

 ing in quantitative inheritance, because it may have a direct bearing 

 on the complex problem of the inheritance of purely quantitative 

 variations in man. This study concerns the amount of red pigment 

 present in the sepals of the flower buds of Oenothera ruhrinervis, a 

 mutant which is characterized in part by having red streaks on its 

 sepals. The extent of this color pattern shows a wide range of 

 variability on either side of the commonest or modal condition. And 

 I have already found that in some cases, at least, the offspring of 

 an individual show the same amount or extent of pigment as their 

 parent form, i. e., they breed true to this purely quantitative differ- 

 ence. This shows that some initial difference in the germ cells de- 

 termines their capacity for producing a certain amount or extent of 

 pigment in the adult individual developed from them under given 

 conditions. 



Some recent work seems to indicate that the different coat 



