414 BOTANICAL GAZETTE [MAY 
are not at all individual in the number of seeds developing per 
locule in the fruits which they produce. The cross correlation 
coefficients for the number of seeds in one fruit and the length of 
another fruit of an inflorescence furnishes no indication that there 
are innate or environmental peculiarities of inflorescences which 
tend to influence both the number of seeds developing and the 
length of the fruit in the same direction. In short, one cannot 
explain the correlation between number of seeds and length as the 
result of superior innate vigor or favorable nutrition in some 
inflorescences of an individual and the contrary conditions in 
others. 
III. From the immediately foregoing considerations, and from 
others detailed in the body of the paper, we seem to be justified 
in the conclusion that the measurable interdependence between 
the number of seeds and the length of the fruit in Sfaphylea and 
probably also in Cercis is a direct physiological one, and that the 
two characters stand in some degree in the relationship to each 
other of cause and effect. 
While this conclusion has already been reached by some other 
biologists depending upon more general evidence, I believe that 
this and the preceding study are the first in which a fairly satis- 
factory approximation to proof has been attained. A chief value 
of these studies is that the numerous difficulties surrounding the 
problem have been more clearly realized than appears to have 
been done before. The numerical results, while substantiating in 
a satisfactory manner the conclusions drawn from them, must be — 
looked upon as merely approximations. 
CARNEGIE STATION FOR EXPERIMENTAL EVOLUTION 
Cotp Sprinc Harsor, N.Y 
