PROGRESS IN BIOLOGICAL INQUIRIES, 1926 651 
early stage of development—in fact, before the embryo leaves the egg. Any 
intermingling between the herring of different localities after these characters 
became fixed would cause all of these differences to vanish. The question 
might arise as to whether each locality was contributing to the others 
through the drift of eggs during the early embryonic period, before the char- 
acters had become fixed. The eggs themselves, however, are attached securely 
to the seaweed and eel grass, and so are not subjected to drift. These facts 
would seem to preclude the possibility that the population of one region con- 
tributes very extensively to that of another, and to show that the fishery is 
uct drawing upon any great offshore body of herring. 
The problem now is to refine the analysis to a point where it can be deter- 
mined if differences occur between closely adjacent localities. This will ne- 
cessitate a careful examination of larger numbers in order to arrive closer to 
the true meaning of the distribution of each character, and to obtain an accu- 
rate measure of the variation to be found between samples. 
Supposing that, through a careful study of these characters, the differences 
give indications of the relative independence of the different grounds; the 
question still remains as to the reason for the changes in abundance. Are 
these changes natural or due to overfishing? 
Great natural fluctuations in abundance occur in both the European herring 
and the California sardine, due to the presence of dominant age groups. Such 
a phenomenon may be one cause of the great fluctuations in the abundance of 
the Alaska herring. To determine this point, samples are being taken for 
length-frequency studies, and the ages are read from the scales. 
The length-frequency distributions of herring taken in July in Prince William 
Sound for the past three years have been studied. These studies show a for- 
ward progression of the modes that may be caused by dominant age groups. 
This is also suggested by the last year’s collections, in which the 5-year-olds 
are subordinate to both the 4 and 6 year olds. 
In order to make certain as to this point, it will be necessary to collect 
data for a few years and note the changes in the position of the size modes. 
The collection of data of this sort presents some difficulties. There is question 
as to the size of samples and the frequency with which they must be taken 
in order to follow the changes in the run. As soon as the opportunity offers 
of obtaining samples for a complete season in one locality, a study will be 
made to determine how often samples must be taken and how large they must 
be in order to secure an adequate representation of the population. 
If regulation is necessary, either because of overfishing or reckless use, 
there must be, necessarily, some knowledge on which to base it. What condi- 
tions should the fish be in when used? At what time do they spawn, and when 
do they reach a suitable condition for packing? At what age do the herring 
spawn? These and many similar problems are being investigated. 
The only data obtained on the age at maturity were taken from beach- 
seined samples at Halibut Cove. No 2-year-olds were mature; out of twenty- 
five 3-year-olds, 18, or 52 per cent, were mature; out of twenty-four 4-year-olds, 
20 or 83 per cent, were mature; all of the 5-year-olds were mature. 
Weights were taken to follow the changes in the weight of herring of any 
given length, as an index to the condition in the various localities and at 
different seasons. Owing to the unusual scarcity of herring it was not possible 
to follow this throughout the season in one locality. The results show princi- 
pally the need of data throughout the season in one locailty, but they have 
some value. 
In order to have a standard of comparison, we have shown the relation 
between the weight and the iength by using a condition factor obtained by 
the formula—K equals one hundred times the weight, divided by the cube 
of the length. It is expected that besides a difference throughout the season 
and between localities, there will also be an annual variation, not only in the 
time when a certain condition is reached but probably also in the maximum 
condition attained. 
Taking herring of 255 millimeters body length a couple of weeks previous to 
spawning, at Halibut Cove in April, when the herring contained no fat but about 
25 per cent of their weight was contained in the ripe gonads, the condition 
factor was 1.35. During the period from June 25 to July 2, in Prince William 
Sound at Elrington Pass, the smaller herring were in good condition and con- 
tained an abundance of belly fat, but the larger herring in the same samples 
contained very little fat and were rather thin, the condition factor at 255 
millimeters body length being then 1.42. In the female gonads from one to half 
