Chap. I. 



PEIMULA VEBIS, 



27 



Eeturning to Table 7, if we consider only the good 

 capsules, those from the two legitimate unions were to 

 those from the two illegitimate in number as 71 to 31, 

 or as 100 to 44. Again, if we take an equal number of 

 capsules, whether good or bad, from the legitimately 

 and illegitimately fertilised flowers, we find that the 

 former contained seeds by weight compared with the 

 latter as 50 to 24, or as 100 to 48 ; but if all the 



Fig. 2. 



Legitimate union. 

 Complete fertility. 



poor capsules are rejected, of which many were pro- 

 duced by the illegitimately fertilised flowers, the propor- 

 tion is 54 to 35, or as 100 to 65. In this and all other 

 cases, the relative fertility of the two kinds of union 

 can, I think, be judged of more truly by the average 

 number of seeds per capsule than by the proportion of 

 flowers which yield capsules. The two methods might 



