540 Harris: OVULES AND SEEDS IN CERCIS 
1 — .60? bass 
CG, = _  —— =. F 
V 100 . 
This would have its probable error 
67449 X .0640/V 220 = .0029.* 
By comparison we get: 
Empirical, S. D. = .1165 + .0053, 
Theoretical, S. D. = .0640 + .0029, 
Difference = .0515 + .0060. 
The difference is over 8 times its probable error and I think indi- 
cates that there are real biological differences in the individuals. 
The constants for the correlation between the number of ovules 
per pod and the deviation of the number of seeds from their 
probable value, 7,., for the 60 trees with 100 pods per tree and 
for the 40 other trees with less than 100 pods are summarized 
from Table V in Table VII. For a grand total of 110 trees I add 
TABLE VII 
FREQUENCIES OF VALUES OF 792 FOR SERIES OF INDIVIDUALS 
Oven md Copsey, «| ET | gece | roel no Tre 
—.325 — —.275 | 2 | — 2 
—.275 — —.225 3 | t 4 
“gs 87S 3 | 5 9 
175 — —-125 3 | 2 7 
—.125——.075 10 | 4 14 
~.075 — —.025 | II | 6 20 
25—+.025 8 | ; 16 
+.025 — +.075 4 | 6 te 
075 — +.125 8 5 15 
+.125 — +.175 | 8 a 
+.175 — +.225 | a | I | I 
+.425 — +.275 | ~ | i | I 
+.275 — +.325 — | 2 2 
| 
the constants obtained for the 10 individuals from which all the 
pods were taken. 
These results are also shown graphically in Fig. 2. Here the 
length of the lines indicates the magnitude of the correlation, 
i. e., the amount by which it deviates from o, and the nature and 
* The 10 trees yielding more than 100 pods each have been included. 
