510 WHITE— STUDIES OF INHERITANCE IN PISUM. 



by crossing them with various F^ segregates breeding true to gray 

 seed coat color. 



In F5, from 60 crosses of colorless F^ X gray F^, 



9 crosses gave 21 gray purple dotted 123 gray, 

 23 crosses gave only gray purple dotted, 

 28 crosses gave only grays without purple dots. 



Tschermak (86) has made numerous crosses between pure varie- 

 ties and extracted F^, F3 and F^ segregates with and without the 

 character purple dotting. In these crosses, colorless X gray without 

 purple dotting in some cases gave all purple dot progeny in F^ 

 (agreeing with the results of Lock's crosses above). In other cases, 

 using different varieties, Tschermak always secured only non-purple- 

 dotted progeny both in F^ and in Fg, except in certain very excep- 

 tional cases. In these exceptional cases purple dotting appeared 

 sporadically on the seed coats of gray segregates which had bred 

 true to a self gray for several generations, while on the other hand 

 there were cases in which purple dotting was expected, but failed 

 to appear when certain crosses were made (86, S. 160). Varieties 

 (86) practically breeding true to the absence of purple dotting also 

 occasionally have a few seeds with purple dots, and these appear 

 on plants the majority of the seed of which is without the purple 

 dots. 



Colorless X orange-brown or greenish orange tinted (e. g., P. 

 arv. Svalof No. VI.) gave in F^, in Tschermak's experiments (86) 

 progeny with dark brown seed coats with purple dots. In F^, 4 

 classes appeared — dark brown with purplish reddish dots, dark 

 brown with no dots, whitish brown (gray) with no dots, colorless. 

 The numbers were small, hence the ratios are not of much im- 

 portance, except in showing that the dark browns were in greater 

 number than the other two classes. The gray segregates were con- 

 stant and in back-crosses with the colorless seed coat parent gave 

 only dark browns, grays and colorless seed coat segregates, with or 

 without purple dots, as in Lock's crosses of F^ colorless and F4 

 gray seed coat segregates given above. If large enough numbers 

 had been obtained Tschermak (86, S. 161) believes the orange- 

 tinted grandparental type would have appeared again. 



Correns (14) crossed a colorless seed coat variety with two 



