Ernest Warren 
135 
was calculated and a correlation table was made between it and the percentage 
death-rate, the coefficient is so small (no. 7) and the probable error so great that 
we can only say that perhaps there was a slight tendency for families with high 
ratio to have a small death-rate. We may assume that adult families which 
exhibited considerable divergence were parts of families, which if all the members 
had survived would have shown the same or greater divergence, hence we are 
justified in the assumption that the deaths which occurred were not at all, or very 
slightly selective with respect to the dimensions with which we are here concerned. 
TABLE II A. 
Number 
X 
y 
1 
Length of Antenna of Mother 
Number in newly born Brood ... 
•120 + -086 
2 
Frontal Breadth of Mother ... 
Number in adult Brood ... 
•129-t--086 
S 
Ratio of Mother 
)) )) )) ••• 
•119 + -086 
h 
Length of Antenna of Mother 
Percentage Death-Rate of Family 
- -201 + -084 
5 
Frontal Breadth of Mother... 
iy n J) 11 
-•184+ -084 
G 
Ratio of Mother 
IT 1? »i )? 
- ^070 + -087 
7 
Mean ratio of Family 
71 11 V 11 
- •113 + -086 
8 
Number in newly born Brood 
11 11 11 11 
-•188 + -084 
9 
Difference between the rati(j 
of Mother and mean ratio 
of Family ... 
11 11 11 11 
-•128 + •086 
10 
Difference between the ratio 
of Mother and mean ratio 
of Family ... 
Number in newly born Brood . . . 
- •OO? ± -087 
Further, the fertility of the mother, measured by the number in the newly 
born brood, was apparently slightly correlated with the death-rate (no. 8), that is 
offspring of fertile mothers tended to be healthier than those of less fertile 
mothers. 
Also, 'families in which the mean ratio diverged widely from the ratio of the 
parent tended to be healthier (no. 9) than those families which more nearly 
resembled the parent ; but there was apparently no correlation between this 
resemblance and the fertility of the mother (no. 10). 
(6) Organic Correlations. 
For any consideration of the correlation of one organ of a parent with some 
other organ of the offspring it would be interesting to ascertain the inter-relations 
of the parts in the individual. The correlation table for frontal breadth {AB) and 
length of antenna (OD) of the second generation is given in Table III. The 
standard deviation of ^5 — 1'3415 and of CD = 2*4744 ; the coeffieient of corre- 
lation is "8023 + "0125. Hence the two dimensions are closely correlated. 
