[THOMPSON] RIPENING PERIODS IN WHEAT 71 



portant Mendelian factor is concerned. But in F3 he found four 

 more or less well marked groups. He therefore assumes the existence 

 of two factors, a chief one (A) and a subsidiary (B). B makes the 

 flowering time later than it would be if A alone were present. He 

 further assumes that because extracted early and late plants were 

 neither so extreme as the parents, "gametic contamination" has 

 occurred. 



Castle (1916) has reviewed Hoshino's results and given a different 

 interpretation. He maintains that the existence of only one pair of 

 factors along with the occurrence of gametic contamination need be 

 assumed. The results as a whole he uses in a criticism of the multiple 

 determiner hypothesis of blending inheritance. 



Leake (1911) has made a detailed study of blooming time in 

 cotton. The Fi plants were again uniformly intermediate and the 

 F2 showed a much increased range of variation. The frequencies in 

 F2 formed a regular curve with mode and mean nearer the late parent. 

 The further fact was established that early flowering is correlated 

 with monopodal branching and late flowering with the sympodal type. 



3. Materials. 



In this investigation eight varieties of wheat were used. One 

 (Prelude) is the earliest valuable wheat which we have. One (Club) 

 is the latest which will mature in our season. The others represent 

 various stages between these two extremes. The plants of each variety 

 used in crossing were grown in pure lines for several years and proven 

 to be stable, except for environmental variations, in regard to the 

 character in question. 



Table I gives the length of the ripening period in 1917 for each 

 of the varieties used. It will be observed that the variation in each 

 pure line extends over a period of about eight days and that the 

 frequency distribution is that of a regular probability curve. The 

 difference between the mean of one variety and that of the variety 

 nearest to it varies from 1 to 7 days, and the difference between the 

 means of the earliest and latest is 23 days. The difference between 

 the earliest plants of the earliest variety and the latest plants of the 

 *latest variety is 29 days. 



It is admitted that the' term date of ripening is used very loosely 

 in practice. The date at which different observers consider a plant 

 to be ripe varies over several days. The reason for this is that differ- 

 ent men emphasize different points — color of straw, color of glumes, 

 dryness of straw, condition of endosperm, etc. In order to avoid 

 this difficulty and to make the results as accurate as possible it was 



