IX 



HEREDITY 



167 



The weight of the seed. The inheritance of the 

 mean seed-weight is particularly interesting. In the first 

 place it fluctuates more than any other character, except- 

 ing the height, and it further shows clear evidence of 

 autogenous fluctuation. 



The first cross in which this character was carefully 

 examined was Afifi x Truitt (Fig. 66), where the mean 

 seed-weights were 0105 g. and 0135 g. The seed-weight 

 in F! was 0165 g. In F 2 the weights ranged from 08 

 to 0175 g., with two marked modes at 0'095 and 0115. 

 The form of this F 2 graph suggested that light seed was 



P-E. 



100 -150 



FIG. 66. MEAN SEED-WEIGHT. 

 Afifi x Truitt, showing extraction of small seed in F 3 . 



segregating from heavy seed, and on testing this by breed- 

 ing on, we found no reason to modify this conclusion ; 

 thus, a plant with seed weighing 0100 regressed slightly 

 in F 3 to a mean of 0'090, with a scatter from O'OYO to 

 0110, and no higher probable error than the parent strain ; 

 and F 4 raised from a 0'090 plant of these gave the same 

 result, ranging from 0'065 to 0110. The plants of the 

 F 4 had been extremely diversified in most other characters, 

 such as height, while the F 4 was almost a pure strain. 



Whether the segregation was simple or compound, it 

 was clear that the size of the seed expressed by us as 

 weight was an inherited characteristic. 



In another cross, namely Charara x King (Fig. 68), the 

 matter became more interesting, and the inheritance 



