J. A. Harris 
481 
irregular as our asymmetry grades, has led me to omit the probable errors for the 
individuals. The results for Series A and C are shown graphically in Diagram 4. 
Sei-ial Ntimber of Individual. 
DiAGEAM 4. Showing mean differeuce in asymmetry of eliminated and matured ovaries for the 
28 individuals and for the population. Circles = mean asymmetry of eliminated ovaries for 
individuals; solid dots = mean asymmetry for matured ovaries for individuals. Broken transverse 
line = mean asymmetry of all eliminated ovaries; solid transverse line = mean asymmetry of all 
matured ovaries. 
The usual comparisons (B — A, C — B, C — A) are given in Table XVIII. 
The differences are taken in such a way that if selection tends to eliminate the 
less symmetrical pods the differences will have the negative sign. For the 
differences we find : 
B — A, 12 positive and 16 negative, 
C — B, 6 positive and 22 negative, 
C — A, 7 positive and 21 negative. 
The mean difference per individual is : 
B— A = - -0064, 
C— B = - 0075, 
C— A = - 0801. 
