A. St Clair Caporn 257 



5. The ^2 crop was harvested all at once. Hence no definite data 

 were available. Evidence of early, late, and many intermediate forms, 

 however, was visible. 



6. Systematic, gradual harvesting of the F^ generation was under- 

 taken, together with parent rows grown under exactly the same 

 conditions, to determine the constitution of the individual F^ plants. 



7. Analysis of results. Earliness is possibly a function of 3 factors. 

 A type which is comparatively early, in that its -^3 period never extends 

 into the period of the late parent, is segregated on a 1 : 3 basis. 



8. Suggestions are made with regard to the improvement of early 

 varieties. 



Since completing this paper the writer has been referred to the 

 work of Hoshino (1915) on peas and rice. In the case of the latter 

 this investigator made observations on the " shooting times " in a cross 

 between an early and a late variety. By " shooting time " is meant 

 the time at which the first spikelet of the inflorescence emerges above 

 the ligule of the sheathing leaf Apparently in rice this is also an 

 exact measure of the relative maturing time of the grain later on. 

 Hoshino's F3 results were remarkably similar to those given in this 

 paper for a different cereal examined on a different system. Two only 

 out of forty-six randomly taken F3 rows were as early as the early 

 parent, while twenty-one stretched across the "gap-period," but not 

 into that of the late parent. As only ten plants per row were raised, 

 the number twenty-one is probably too high, and would, had there 

 been more plants, have been nearer 11 — 12, which is the expectation 

 on the 1 : 3 ratio. Although he overlooks this important fact, however, 

 Hoshino nevertheless strongly favours the multiple-factor hypothesis 

 as an explanation, deducing from the very small number of pure eaflies 

 like the original early parent that there are three factors. 



The value of this present paper, therefore, does not lie in the 

 originality of its inferences, but rather in that similar conclusions to 

 those of the Japanese worker have been reached perfectly independently 

 by different methods, and also in a cereal. 



LITERATURE. 



Hoshino. 1915. Imperial Unioersity^ Sopporo, Japan. Vol. vi. November. 



