50 — 
and compressed, head and tail however, being long (long tailed fish). They may be 
divided into several minor groups: 
a) Coast herring of the northern waters of the North Sea and the Skagerak. Ex- 
hibit considerable resemblance to the Norwegian spring herring. 
b) Coast herring of the southern waters of the North Sea, the Kattegat, and the 
Western Baltic. These are smaller, (less than 25 cm.) with fewer vertebrae. 
c) Spring herring from Rügen. 
3) Ocean herring of the North Sea Banks, inhabiting the open sea from the coasts 
of England and Scotland across the whole of the North Sea, through the Skagerak and 
Kattegat, and in to the Western Baltic. In summer and autumn they move to seek 
spawning grounds on the sandy and stony banks which rise from the depths of the sea 
at some distance from land. 
These have all a medium number of vertebrae, (56.5—55.5) a large number of keel 
scales behind the ventral fins (15—14 average) these scales being highly developed. They 
are broad skulled fish, with long body but short tail (short tailed fish). 
a) The northern bank herring belong to the northern waters of the North Sea, Skage- 
rak, and Kattegat. To these must be reckoned first of all the great race of 
herring which forms the object of the Shetland, Scottish, and North of Eng- 
land fishery; in addition, also those herring which spawn on the Jutland 
Bank and are taken in great numbers in winter off the coast of Bohuslan. 
No. of vertebrae 56.5, body long, head, and especially tail, short. 
b) The southern bank herring. Closely related to the foregoing, differmg essentially 
by a larger number of keel scales between the ventral fin and anal aperture, 
and greater breadth of head. The group includes the herring south of the 
Dogger Bank. 
This system represents in many ways a very high degree of progress, based as it 
is upon investigations carried out according to a clear and critical method, from which 
definite arithmetical values may be obtained for distinction between the different forms. 
Mixture of Races in the North Sea. 
We have seen in the foregoing, that HEINCKE had arrived at a clear understanding 
of these races or types, by investigation of a large number of individuals, the average 
characteristics of which presented a distinct view of the racial peculiarities. It is however, 
obvious, that such a method of investigation must presuppose the samples as “pure”, 
i. e. consisting of individuals of the same race. If samples containing a mixture of speci- 
mens belonging to two or more races are treated statistically as one, then the results 
obtained will naturally be valueless or misleading. Now the different races in the North 
Sea frequently occur in unmixed shoals; they are, however, also often found mingled 
together. Samples of such shoals will then also be mixed, and only by sorting the samples 
will it be possible to separate the different races one from another for individual examina- 
tion. Such sorting can also be carried out according to Hrıncke’s method; one can, 
as HEINCKE points out, instead of examining a large number of individuals, examine a 
large number of features in some few specimens, or even in a single one. Such investi- 
