1912] HARRIS—STAPHYLEA 413 
fruit. If the size of the fruit increases as the number of seeds 
becomes larger, the development of the seed must exert a stimulus 
to the development of the fruit wall, providing that the correlation 
between the number of seeds and the size of the fruit is not due to 
some other factor or factors upon which both seéd number and 
fruit size are in some degree dependent. 
The task is, therefore, twofold: (a) to obtain a measure of the 
correlation between the number of seeds and the length of the 
fruit, and (6) to show by a process of elimination that the corre- 
lation can, with a high degree of probability, be attributed to a 
direct ore relationship between number of seeds and size 
of fru 
i; ce first of these undertakings is straightforward. The 
coefficients of correlation show that in both Cercis and Staphylea 
there is a very substantial interdependence between number of 
seeds and fruit length. 
II. The second task is somewhat more complicated. The 
following facts indicate that this observed interdependence is due 
to physiological factors confined to the seed and ovary wall: 
(1) In Staphylea the correlation between the total number of 
seeds and the length of the fruit is higher than that between the 
number per locule and length. 
(2) The relationship between the number of seeds developing 
and the length of the pod is in large measure independent of the 
influence of the number of ovules. 
(3) In both Cercis and Staphylea the possibility of a mechanical 
stretching of the fruit through the pressure of adjoining seeds seems 
to be excluded. 
(4) Both length of pod and number of seeds developing are 
slightly correlated with the number of fruits per inflorescence and 
with the distance of the node from the base of the inflorescence, 
but the correlations are too low to be of any significance in pro- 
ducing the relationship between the number of seeds and length 
ot pod. 
(5) The inflorescences of a shrub of Staphylea seem to be slightly 
differentiated with respect to the number of ovules per locule and 
the length attained by the fruit. Apparently the inflorescences 
