408 BOTANICAL GAZETTE [way 
TABLE XXI 
SUMMARY OF INTRA-INFLORESCENCE CORRELATION COEFFICIENTS 
LENGTH AND 
VULES SEEDS 
OVULES AND SEEDS AND LENGTH AND LENGTH AND 
ULES _ | SEEDS LENGTH oO 
f Average f Average f Average f Average 
Positive...| 19 |+-o.149 | 12 +0.031 16 |+0.146 | 14 |+0.077 | 10 |+0.062 
Negative..| -1 |—0.026 | 8 |—0.022 | 4 |—0.023 | 6 |—0.133 | 10 |—0.037 
Total. ..| 20 |+0.140 | 20 |+0-010 | 20 |-++o.112 | 20 |+0.014 | 20 |+0.025 
appear in table XX, and these are still further summarized in 
table XXI. 
The intra-inflorescence correlation for ovules is unquestionably 
positive, and perhaps high enough that one safely can say that the 
inflorescences are differentiated among themselves with respect to 
number of ovules produced. For number of seeds per locule the 
inflorescences seem not at all differentiated; it is not even possible 
to ascertain the sign of 7,,”, and its mean is only o.o10. This 
result seems to me of considerable physiological interest. I had 
expected to find a greater similarity among the numbers of seeds 
developing in the locules of the inflorescence than among the 
numbers of ovules formed. I thought that probably some inflores- 
cences would be in much more advantageous positions for obtaining 
_ food material than others, and that in consequence the differences 
between inflorescences would be greater (and consequently the 
intra-inflorescence correlation higher) than for number of ovules. 
So far as we can judge from evidence at hand, the morphogenetic 
factor is stronger than the physiological? in determining the charac- 
teristics of the fruits of an inflorescence. For length, it seems rea- 
sonably certain that there is a slight similarity between the pods 
9 One must use extreme caution in such fields as this, for there are innumerable 
pitfalls. By morphogenetic I mean the organogenetic processes which give rise to 
the ovaries. By physiological ae gh I refer to (a) the ecological factors which deter- 
mine whether an ovule shall receiv sperm, (b) to the availability of food material 
and other requisites for growth, @ ot innate vigor of the individual aPeapr 
which determine whether a fertilized ovule shall develop into a seed. The third of 
these is probably in some measure identical with the organogenetic. If (a) a (db) 
were really very different for the several inflorescences of an individual, one would 
expect the intra-inflorescence correlation for number of seeds to have sensible values 
and possibly to rise considerably above what it is for ovules. This is the case in 
Sanguinaria (Biometrika 7:328. 1910), where it seems that the main bulk of the 
