202 



GENETICS AND EUGENICS 



ing, F2 was intermediate but highly variable, covering prac- 

 tically the entire range from the flowering time of the early 

 to that of the late parent. F3 was also highly variable but a 

 few families were found to be as ''constant " in flowering time 

 as the parent varieties, and in r4 the proportion of constant 

 families had increased further. Two hundred and thirty of 



Fig. 132. A cross of two varieties of peppers differing greatly in size and shape of fruits. Fruits of 

 the parent varieties are shown at P and P, of Fi between them, and of F2 in the four lower rows. 

 Each fruit is taken from a diflFerent plant and is typical for the plant. (After Gross.) 



the four hundred and twenty-one F4 famihes studied by 

 Hoshino were found to be as "constant" in flowering time 

 as the parent varieties. The mean flowering time in days 

 from sprouting as observed by Hoshino is given in Table 32. 

 It will be observed that the white-flowered F4 constant 

 families were all early or intermediate in flowering time 

 whereas the red-flowered families were chiefly late. This 

 clearly indicates linkage, or coupling, between flower color 

 and time of flowering. But flower color clearly Mendelizes, 

 hence flowering time must also depend upon a Mendelizing 

 gene, which is linked with the gene for red flower color. 



