— 204 — 
great reductions taking place in the numerical value of the year classes at later periods 
of life. All that it is desired to assert is that the observations here recorded distinctly 
point to the conclusion that the fluctuations noted in the stock have their origin in cer- 
tain conditions prevailing at a very early period in the life of the fish. 
Extent of spawning and numerical value of year classes. 
The first question which then arises is whether the numerical value of the year classes 
is dependent upon the extent of the spawning; i. e., the number of eggs produced. As 
mentioned in the previous chapter, this question has never yet been investigated. It 
would also be a task of extreme difficulty, especially in Norwegian waters, on account 
of the irregular distribution of the eggs (vide Chap. III) which renders it impossible to 
accurately ascertain the entire quantitative value of the eggs, at any rate, with the means 
and methods at present available. Little can therefore be stated as yet beyond the 
results indicated by the comparison made in the previous chapter between the quantity 
of roe observed during the fishery, and the number of fish in subsequent years. It will 
be remembered, that this comparison showed the very years in which the quantity of roe 
at Lofoten was least to be those which produced the richest year classes. This being so, it 
is difficult to avoid the conclusion that the actual quantity of eggs spawned is not a factor 
in itself sufficient to determine the numerical value of a year class. A rich spawning 
may produce a year class poor in numbers, while a large year class may have its origin 
in a year when the spawning was at its lowest. 
Larvae and young fry stages. 
We must therefore look to the later stages of the eggs to find the conditions which 
determine the numbers of individuals in any year class. This again leads us to the que- 
stion, at which stage of developement the most critical period is to be sought. Nothing 
is known with certainty as to this; such data as are available, however, appear to indicate 
the very earliest larval and young fry stages as most important. 
In the first place, we have the significant fact that several different species of spring 
spawning fish (cod, haddock, and Norwegian spring herring) all exhibit remarkably 
high values for one and the same year class, (that of 1904) whereas the North Sea herring, 
for instance, which are summer spawning, show specially high figures for the fish of quite 
other years. This would seem to indicate that the spring of 1904 was marked by the 
prevalence of certain hydrographical or biological conditions which determined the 
production of this year, and as the herring, cod, and haddock of this year class were 
at that time in the larva or young fry stage, it is natural to devote especial attention 
to the consideration of these stages. Two points immediately suggest themselves in 
this connection, viz; the conditions as regards nourishment to which the fish were subject 
at this stage, and the passive movement of the same stages under the influence of the 
currents. 
Nourishment of the larvae. 
It has long been a recognised fact, both in connection with artificial meubation 
of fish eggs and in the study of the youngest egg and larval stages in the sea, that the 
numbers of individuals rapidly decrease as soon as the newly escaped larvae have consu- 
