} 
: 
{ 
: 
| 
HARRIS: RELATIONSHIP OF OVULES TO SEEDS 249 
ing their ovules into seeds as those which do not produce so many, 
but that the relationship is a very slight one. 
The differences between two random samples taken so closely 
together and the very low correlation makes one cautious in 
accepting our constants as biologically significant for Cercis as 
a species, or even for Cercis as a race growing at Meramec 
Highlands. Under the circumstances the only thing to be done 
is to collect wider series of data. 
The collection of this additional material was carried out from 
two standpoints; first that of widening the sources of pods in 
number of trees and variety of habitats, second, that of securing 
greater homogeneity in the series of pods upon which individual 
constants are based by taking them all from individual trees. 
The discussion of the results of analysis of data for the individual 
trees must be reserved for a later contribution. In addition to 
the general samples just described from Meramec Highlands, 
smaller lots were taken from about 125 trees. These can be 
added to the 10,000 pods already discussed. 
TABLE V 
SEEDS 
Ovules eC Peas 3 4 5 6 7 8 
I 7 — — _ on apes 7 
2 baer Be Corse aioe Meme eigae! 348 
3 | 190 837 | 1,407 | Sloe et ee 2,434 
4 | 236 | 1,264 | 3,536 | 4.582 | — ee eons are 9,618 
5 100 | 620! 1,691 | 3,630 | 3,955 Secret, ace 9.996 
6 46 | 142| 444 | 1,064 | 1,787 | 1,806 | lbes 5,289 
7 2 21 42 { SERO4 GOL ft Rge oF a 826 
8 _— —_— Toe Tee es Oe 36 
648 | 3,166 | 7,120 | 9,368 | 5,932 | 2,073 | 242 5 28,554 
Consider now the total material, amounting to 28,554 pods, 
from Meramec Highlands. The data appear in Table V. The 
constants are: 
A, = 4.7020 + 0040, A, = 3.8399 = .0048, 
% = 1.0074 = .0028, Od, = 1.2036 + .0034, 
Vo, = 21.425, Ve. = 31.3452. 
Yor = .6482 = .0023, for = — .0463 * .0040. 
