110 Glume-Length and Grain- Length in Wheat 



of glume had revealed the segregation of " long " and " short " glumes 

 (apparently of the parental forms) in the F^ and, in addition, of hetero- 

 zygous forms of intermediate length. It seemed desirable to determine 

 whether or not the numerical limits which characterised the parents 

 of the cross, also characterised the apparently " parental " types found 

 in the F^. 



Glume-length was chosen as the experimental character partly 

 because of the general definiteness of BifFen's results, and further, on 

 account of its botanical suitability to the requirements of a simply 

 measurable character. It was such that numerous and accurate measure- 

 ments were obtainable, and moreover, as a general experience proved, 

 it was less liable to wide and sudden fluctuations than were most of the 

 other length characters of the wheat-plant. 



Grain-length it was clear, bore some definite relation to glume- 

 length [vide (1), p. 38], and as it appeared to have an economic signifi- 

 cance, it was included as an experimental character in the investigation. 



The parents {Fq) of the cross were " Polish " {T. polonicum L.) and 

 "Kubanka" (a variety of T. durum Desf). They had been grown for 

 some years by Professor Biffen, and were known to satisfy the require- 

 ments of a definition of " pure line " with regard to origin, behaviour, 

 and mode of maintenance. 



The sequence of the breeding was : 



f Cross made, viz. $ Polish x (^ Kubanka. 

 ( Parental stocks continued. 



[ ^1 plants grown and harvested. 

 I Parental stocks continued. 



' Every grain borne by the Fi plants was sown and 530 F^ 

 plants thus raised. The main ear of every F^ plant was 

 1913 \ harvested and labelled with a serial number (1 — 530). Of 

 these plants only 487 set ripe grain. 

 \ Parental stocks continued. 



' Every grain of every labelled ripe Fz ear (one per plant) was 

 sown, and 5145 Fg plants were thus raised. The main ear 

 of every plant was harvested. The ears from those plants 



; which sprang from the seeds of any one F^ plant were 

 bunched together, and the bunch was labelled with a 

 "family number" which was the number allotted in 1913 

 to the corresponding F2 parent plant. 

 Parental stocks continued. 



1911 



