[THOMPSON] RIPENING PERIODS IN WHEAT 75 



similar to those recorded for the 1916 and 1917 seasons have been made 

 over a series of years. 



All the data given in the following pages for the parental varieties 

 and F2 refer to the season of 1917. The data on the Fi plants were 

 collected in 1916 and due allowance made for the difference between 

 the two seasons. 



4. Experimental Data. 



Crosses were made between many different pairs of the eight 

 paternal varieties. The whole work was planned so that data could 

 be secured on crosses between parents showing all possible degrees of 

 difference in regard to the characters in question. Furthermore, 

 crosses involving parents differing only slightly were made over the 

 whole range of variation. In this way it was hoped that conclusions 

 which might be drawn in regard to hereditary differences between 

 extreme varieties by means of direct crosses could be checked by con- 

 clusions drawn from adding up the differences between intermediate 

 points. 



(i) Ripening Period. 



Some of the results in regard to the length of the ripening period 

 are given in tables IV to XIII. These tables are arranged in the order 

 of the earlier parents. Thus tables IV to VIII include the crosses 

 of the earliest (Prelude) with successively later parents; IX, X, XI, 

 include crosses of Bobs (second earliest) with successively later 

 parents; XII gives the results of crossing Marquis (third earliest) with 

 a later variety; and XIII, of crossing Red Fife (fifth earliest) with the 

 latest of all. 



In order to make comparison easier, each table gives in succes- 

 sive lines (1) the number of days from planting to harvest, (2) the 

 range of variation of earliest parent, (3) of latest parent, (4) of Fi, 

 (5) of F2. 



The First Hybrid Generation. — It will be observed that in each 

 case the first hybrid generation ripened with the later parent. This 

 result is obtained not only by making due allowance for the difference 

 between the 1916 and 1917 seasons but also by actual comparisons 

 with the paternal varieties in 1916 (when the Fi data were collected). 

 The numbers of the first hybrid generation plants are of course not 

 large enough to reveal the extremes of variation or the form of the 

 variation curve. Consequently one cannot state whether the Fi 

 variation coincides with that of the later parent. But certainly most 

 of the Fi plants appear to be near the mean of this parent. 



