R. N. Salaman and J. W. Lesley 



31 



Totals 



914 



13 



plants occurred. Two of these families arose from a cross of the type 

 " prostrate " x " upright," the third from the reverse cross, and in each 

 case there is a small but definite proportion of " prostrates." Two of 

 the "upright" parents (Kew 19 and Flourball) have been selfed and 

 yielded spreading but no "prostrate" progeny, but of the third, 1/5532 x 

 Fh 26, we know nothing of its decendants. 



Although some of the families are small, yet taken together the figures 



"Upright" 914, 



" Prostrate " 13, 

 show clearly that the " upright " character behaves as a dominant and 

 the " prostrate " as a recessive. 



The incomplete dominance observed in the three F^ families out of 

 the total twenty-three may possibly be due to faulty technique, seeing 

 it has occurred only amongst those crosses which were not protected by 

 bags. On the other hand, there is good reason tcr expect that our 

 "upright" parents are not all of the same genetic composition as regards 

 the factors for " prostrate," and in this relation it is particularly inter- 



