1924] 
HARRIS—THE INFLORESCENCE OF MANFREDA VIRGINICA 453 
available that one of these relationships is more intimate than 
the other. 
The low correlations between the actual position and the rel- 
ative position of the fruit and the number of seeds which it 
produces may be due either to an absence of relationship between 
these two characters or it may be due to a periodicity in the in- 
florescence, such that the number of seeds at first increases and 
then decreases as one passes from the proximal toward the distal 
end of the inflorescence. 
To test this matter we determine the mean number of seeds in 
each position, both actual and relative, on the inflorescence, and 
ascertain whether the change in mean number of seeds occurs at a 
uniform rate from the lower to the higher regions of the inflores- 
сепсе. То be strietly valid the correlation coefficients should 
represent cases of linear regression. The linear equations showing 
the change in the number of seeds per locule with position on the 
inflorescence are as follows: 
Actual position Relative position 
Meramec Highlands, 1906 s = 14.5551 — .0519 a 8 = 14.3458 — .1280 r 
Meramec Highlands, 1907 | s = 13.9185 — .0787 a в = 13.3141 — .1279 r 
Jefferson Barracks, 1906 s = 11.6245 — .0980 a в = 11.6010 — .3769 r 
Jefferson Barracks, 1907 &=1 .0850 s = 11.8913 — .2491 r 
Here a = actual position, 7 = relative position, and s = num- 
ber of seeds per locule. 
The lines for actual position are represented in fig. 11, while 
those for relative position appear in fig. 12. For both relation- 
ships and in all series of material these indicate a uniform de- 
crease in number of seeds per inflorescence when the capsules are 
considered in series ranging from the proximal to the distal regions 
of the inflorescence. 
For comparison we have a number of determinations on the 
inflorescence of Staphylea, in which the relationship between 
position and number of ovules (Harris, '11 a) and between posi- 
tion and number of seeds (Harris, ”12 b) have been shown to be 
very small numerically but generally negative in sign. Тһе re- 
lationship between position on the inflorescence and length of 
pod in Staphylea has also been shown (Harris, 12) to be slightly 
negative. 
