26 



Rogueii " /» Culhiart/ Peaa 



The interrelations of classes 4 and 5 are equally obscure. Inasmuch 

 as class 4 never breeds true, it cannot be regarded as homozygous in the 

 usual sense ; nor in view of the fact that it throws almost exclusively 

 plants of class 5, can it be represented in any simple way as hetero- 

 zygous. 



Crosses between Types and Rogues. 



We have next to consider the remarkable evidence provided by the 

 crosses. It relates in the first place to 52 families raised by crossing 

 thoroughly typical plants with rogues of classes 4 or 5. Of those cross- 

 bred families 50, containing 284 plants, when fully firoivii, were 



CROSSES. 



1. Ti/pe 9 X Rogue, (J. 



Offspring 



