GROWTH AND MATURITY OF SALMON IN THE OCEAN 
19 
or of perhaps no growth, which obtains during the fall and winter, the rings are 
crowded together closely and are frequently more or less broken and imperfect. 
The complete year's growth, therefore, consists in a "summer" band of relatively 
wide rings followed by a "winter" band of narrow rings. By counting the num- 
ber of "summer" or "winter" bands the age can readily be determined. 
Weymouth (1923) has given a brief though admirable discussion of the various 
methods of determining age. In answering certain criticisms that have been made 
of the method of age determination of fishes by means of scales, he makes the 
following pertinent observations : 
These objections are not valid, however, in the case of a number of species of fish, where tlie 
annual nature of the marks rests on no assumption of any kind, but on direct observation. In 
some a study of the scales throughout the year has clearly shown that the ring is formed during 
the winter and only once each year. In others it has been shown that the number of rings agrees 
with the known age of fish kept in captivity or of marked fish recaptured after known periods. 
The soundness of these conclusions is not affected by the fact that in certain species the rings 
are less distinct and hence in these cases may be an unreliable guide to age, nor that incompetent 
or hasty workers may have reached incorrect conclusions in any species. 
Although this method of determining the age of fishes appears to be simple 
and of easy application, in practice serious difficulties are frequently encountered. 
The relative approximation of the rings is merely a reflection of the rate of growth 
prevailing at the time the rings were formed. Not infrequently factors cause 
variations in the rate of growth which are unassociated with the seasonal varia- 
tions responsible for the formation of "summer" and "winter" bands. As a 
rule, the cause of these incidental variations is unknown, although the present 
writer has suggested causes for some of the minor variations observed in the scale 
growth of chinook salmon previous to or during their seaward migration (Rich, 
1920). Snyder (1922 and 1923) has also discussed the causes responsible for 
unusual checks on the scales of chinook salmon in the Klamath River. In the 
case of the chinook salmon the determination of age is further complicated by the 
fact that the young fish, at least in the Columbia River, may migrate seaward at 
any time after they emerge from the gravel of the spawning beds up to an age of 
18 months or more. There are, therefore, many difficult and puzzling characters 
in the scale growth which must be worked out before a completely satisfactory 
analysis can be made, but a full consideration of the problems connected with age 
determination in the chinook salmon lies outside the scope of the present paper. 
The age determinations will, therefore, be presented without fiu-ther discussion of 
the details. It may be remarked that the author, in the course of an intensive study 
of the scales of several thousands of chinook salmon from the Columbia River, 
has found nothing that conflicts with the main conclusions reached by Gilbert 
in his initial paper dealing with scale studies (Gilbert, 1913). In all cases where 
the interpretation of the scales was in doubt the individuals have been omitted 
from the tables. This procedure may have eliminated some few small categories, 
but these would not materially affect the main conclusions that have been reached. 
The lack of data bearing on the types of scale growth presented by the fish 
of the various coastal streams and of the various tributaries of the Columbia has 
made the study of the scales of the fish taken in the ocean an extremely difficult 
