378 Darwinism, Biometry and Some Recent Biology. 1. 
greyhounds, Barrington, Lee and Pearson found the ratio of parental to fraternal 
correlations to be -466 to -529, for the horse -522 to -633, for man "460 to '519 ; and 
TABLE V. 
Number of Tentacles of Pairs of " Brethren " (Siblings). 
Number of Tentacles of First Sibling. 
6 
7 
8 
9 
10 
11 
Totals 
5 
23 
11 
3 
42 
a 
1878 
3599 
1426 
231 
28 
7 
7174 
23 
3599 
8968 
4473 
941 
114 
20 
18138 
8 
11 
1426 
4473 
2880 
723 
107 
12 
9632 
9 
3 
231 
941 
723 
254 
63 
7 
2222 
10 
28 
114 
107 
63 
10 
1 
323 
11 
7 
20 
12 
1 
47 
Totals 
42 
7174 
18138 
9632 
2222 
323 
47 
37578 
Harrington and Pearson for shorthorns "40 to '53. Again Warren found for tnean 
results for Hyalopterus and Daphnia "500 to -656. It will accordingly be clear 
that the ratio of fraternal to parental correlation lies between 1*1 and 1'3. If we 
suppose the weakening of heredity to be as great in the filial as the parental 
generation, we should expect fraternal resemblance in Hydra to lie between 
1-1 X -136 and 1-3 x 136, or say between -15 and -18. Our results are: 
(37,578) Members of same Co-fraternity* 
Mean 7 -2 121 
Standard Deviation ... •861) 
Correlation 
•167 + -003 
The correlation of " brethren " is thus significantly greater than the corrected 
correlation, 'ISo, of parent and offspring, the actual ratio being 1'23. It differs 
also significantly from the relation between " uncle " and " nephew," the ratio in 
that case being 1'76, which compares closely with the value for man '50 to '30, 
i.e. about 1'67. It appears impossible therefore, from the three tests made (i.e. 
comparison of the parental correlation with the correlations of individuals with 
their grandparents, with their " uncles or aunts " and with their brethren) not to 
admit that the individual has a special relation to each grade of its line, and 
is not equally related to all members of its line. 
Of the results we have reached that for the foundation stock and their grand- 
children ( 03) is, perhaps, the least satisfactory ; it is so low. But the source of 
this is fairly obvious. Out of each array of offspring of the 26 original Hydra, one, 
two or three, in one case only, four, individuals were selected to carry on the line. 
These individuals were not selected at random, nor were they taken proportionally 
to the frequency distribution of the tentacles in the offspring. The result is that 
Pairs of "brethren " in the first generation. 
