F. L. Engledow 129 



this problem are possible but it has been decided to apply them to a 

 cross whose parents are more widely separated in grain-length, -form, 

 and -weight, than are Polish and Kubanka. 



XI. Conclusions. 



(1) In this cross, length of glume appears to segregate simply and 

 in the F^ the ratio long : intermediate : short = 1:2:1. 



(2) The "longs" and "shorts" of F^ although recognisably of the 

 same form as the parents (^o), yet differ from them in mean glume- 

 length. The average of the longs is reduced by 24*8 °/„ of the ^o value. 

 In the case of the F2 shorts there is an increase over the F^ shorts but 

 it is smaller. 



This change or " shift " in the longs of the F^ is quite definite and 

 cannot be explained by errors of sampling, seasonal variation, or any 

 such cause. 



(3) The "shifted" form of longs of F^ breeds true as far as F^ (it 

 has not yet been carried further). 



(4) When, in the selfing of F^ heterozygotes, plants of parental type 

 are again evolved, there is no evidence of super-added " shift." Likewise 

 there is no evidence of a tendency to the restoration of Fq values of 

 mean glume-length, i.e. 



{F,.P ex F2.I) = {F, . P ex F,.P) = F,. P <F, . P. 



(5) For grain-length results are obtained analogous to (1), (2), (3) 

 and (4). The percentage value of the " shift " is, however, less than it 

 is for glume-length, e.g. 



Fo.P-F,.P = 12'5''/^oiF,.P. 



(6) All the available evidence supports the view that grain-length 

 and glume-length belong to the same generation or, differently expressed, 

 that grain-length is a maternal character. It is difficult to harmonise 

 this view with the double-fertilization hypothesis. 



(7) Numerous grain and glume characters appear to be genetic in- 

 separables. Long-type glumes never contain short-type grains and vice 

 versd. Despite the association of glume-length and grain-length, the 

 correlation between these two variables is low. Certain reasons may be 

 urged in explanation. 



