376 Darwinism, Biometry and Some Recent Biology. I. 
still somewhat selected material as compared with the offspring. Taking, however, 
the regression coefficient as less influenced by such selection, we see that the 
intensity of heredity is essentially the same for all parents as for the foundation 
stock parents, i.e. 135 and '138. 
TABLE II. 
Number of Tentacles of Pai-ent and Offspring. 
Number of Tentacles of Pai-ent. 
S ° 
5 
7 
8 
9 
10 
11 
IS 
Totals 
1 
1 
6 
38 
436 
440 
290 
81 
20 
5 
8 
1318 
86 
862 
875 
724 
254 
81 
20 
2907 
8 
33 
403 
501 
493 
184 
67 
4 
20 
1705 
4 
75 
114 
134 
62 
25 
11 
425 
10 
10 
13 
13 
13 
1 
50 
11 
2 
12 
9 
30 
Totals 
162 
1776 
1942 
1666 
603 
213 
60 
6436 
We have now to consider the next stage, which presents considerable diflficulty, 
for Hanel has not preserved all the offspring in order to reach a random sample of 
the second generation, but preserved two or three from each group only. We took 
TABLE III. 
Number of Tentacles of Stem Grandparent and Offspring. 
Number of Tentacles of Stem Grandparent. 
cles 
5 
10 
11 
12 
Totals 
<s . 
"g bjo 
15 
149 
157 
47 
28 
14 
18 
428 
61 
243 
249 
149 
92 
10 
26 
830 
^ ^ 
8 
44 
116 
121 
124 
66 
6 
22 
499 
9 
14 
32 
29 
22 
15 
115 
10 
4 
2 
1 
? 
8 
11 
4 
z 
4 
Totals 
134 
540 
564 
344 
202 
33 
67 
1884 
however all the " grandchildren " of the original foundation stock with the following 
result : 
Grandparents (2G) Offspring (1884) Correlation 
Mean 7-199 7-181 f -030 
Standard Deviation 1-288 -884 |±-016 
Now on the pure line theory the resemblance of the offspring to the parent 
ought to be no greater than it is to the grandparent. It is here more than four 
times as great. Another very good test of the " pure line," it occurred to me, 
might further be used to iUustrate this point. If the theory of the pure line hold 
for Hydra, the relation of the offspring to the whole fraternity of which the parent 
