﻿'0 Harris: Distribution and correlation of sexes 



lo types of flowers in the inflorescence. This seems a question 

 considerable interest because of the many discussions of the 



fluence of feeding or degree of development on sex. 

 The problem is approached statistically as follows. Let z 









.u.e^ 



OF STAMINATE 1 



LOWERS 

















:'| 



: 





































1 























jl 









-3-0- 



_!_ 









1 ^— ^-^-^ 







:^ 



T 



^ 



\ / 



v 



V 





















11 _ 













. 





. 





t z 





* 





represent the deviation of the number of staminate flowers per 

 inflorescence from the probable number on the assumption that the 

 proportion of the two types, s and p, is unaffected by the total 

 number produced per inflorescence. If the correlation between 



