396 THE WHEAT PLANT 



group inseparable by eye into two classes, but distinct from the short- 

 glumed segregates. The numbers of each were : 



Long and intermediate, 172 ; short, 55 ; or a 3 : i ratio. 



The Polish wheat used in the cross had faintly pubescent glumes, 

 Kubanka being a glabrous-chaffed wheat. The Fj was more pubescent 

 than the polonicum parent. 



In F 2 both smooth and hairy-glumed plants appeared. Among the 

 pure short-glumed plants were three felted to one with smooth glumes : 

 the proportion of pubescent to smooth in the heterozygotes was 85 : 3 1 

 (15 strongly felted), while all the pure long-glumed plants had glabrous 

 chaff, the factor for long glume apparently inhibiting pubescence. 



Results of a cross between the Polish and Kubanka wheats (T. durum) 

 were also investigated by Engledow. 



The F! was intermediate. The glume-length of the F 2 progeny was 

 determined and compared with the glume-length of the grandparents, 

 the measurements being made of a glume from one of the four median 

 spikelets of each ear. 



The mean glume-length of the Polish wheat used in the cross was 

 30-84 mm., that of the durum being 11-7 mm. 



In F 2 three glume-length types appeared, the measurements of the 

 glumes yielding a tri-modal frequency curve, the respective modes being 

 11-5 mm., 16-5 mm., and 25-5 mm. The calculated frequencies of the 

 F 2 segregates were : 



23-29 per cent (Extracted Kubanka), 57-76 per cent (Intermediate), 

 18-94 P er cent (Extracted Polish). 



To determine more accurately the classification of the F 2 , the F 3 genera- 

 tion was raised and measured : the percentages of the three types were : 



23-65 (Kubanka), 55-39 (Intermediate), 20-95 (Polish). 



The figures suggest a i : 2 : i ratio, but deviate considerably from it. 



A comparison of the measurements of the grandparental forms and 

 the extracted polonicums and durums in F 2 showed that the true polonicum 

 and durum types are produced in very small numbers if at all in F 2 , the 

 extracted polonicum having a mean glume-length from 20 to 25 per cent 

 less than that of the original polonicum, and the extracted durums some- 

 what longer glumes than the grandparental Kubanka. 



The reduction or " shift " towards the mean glume-length of the two 

 grandparents is most clearly established in the long-glumed or " poloni- 

 cum " segregates of F 2 , and the " shifted " form breeds true in F 3 , no 

 evidence of additional " shift " or a reappearance of the pure polonicum 

 being obtained. 



Engledow found that in regard to the length of the grain the F x 



