130 TRANSACTIONS LIVERPOOL BIOLOGICAL SOCIETY. 
Thus a comparison of K, in the 383 spring herrmg from 
the Baltic with 254 from the North Sea gives P <-000001, 
pointing to the existence of local races independently of 
the distinction between the autumn and winter herrings. 
As regards my own figures, I have already pointed out 
several reasons why much reliance cannot be placed upon 
them. But I have tried to find if any were sufficiently accurate 
to be employed, with the following results :— 
In the Port Erin fish we have two samples of 100, numbers 
(3) and (4) of 1914 (June and July), which give the following 
results :— 
Ky OP =e. 
D> PHVA 
V2. P=] 0. 
A= P= O18: 
Here we find a close resemblance, justifying us in considering 
these as samples of one population. 
Among the Welsh fish we have two samples of 50 each, 
numbers (8) and (10) of 1914 (December). These are small 
numbers, but the results are as follows :— 
K,:. P= practieally 1. 
DD: P=—F5: 
VY > P=]ps 
A: P=0-48. 
These, again, we may regard as one population. 
Comparing the Port Erin fish with the Welsh, we find :— 
KG. P= As: 
D> P= = oer: 
ae =e 
ee ee 
Thus we can regard it as very improbable that the two 
areas have the same race of fish. Whether this is a case of 
