age at maturity of the pacific coast SALMON 
15 
correctly so interpreted, but representing the first year and a half of the life cycle and 
not the first few months only. Plate vi, figure i, is also a 5-year fish of similar type. 
Plate VII represents, however, a 4- year scale, the second summer showing an “inter- 
calated check” wholly similar to that characterizing the sockeye scale represented on 
plate II. But whereas the sockeye scale presents three winter bands outside the “inter- 
calated check,” the king salmon scale presents but two, hence the necessity for the 
author’s interpretation of the same structure in two opposite ways, in order that both 
may appear 4 years old. As already indicated, the sockeye scale presents five }^ears’ 
growth and the king salmon four. 
We shall not here enter upon a detailed discussion of accessory bands formed by 
checks in growth during the summer, especially during the first summer in the sea. Such 
have been demonstrated by Johnston to occur in the Atlantic salmon and are abundantly 
represented in any series of Pacific salmon or steelhead scales. Their true nature can 
usually be recognized without difficulty — as in the specific cases mentioned — by the 
proportion of the bands in which they occur and often by a wide variation in their 
appearance in different scales from the same fish. Occasionally, however, they so 
closely simulate genuine winter bands as to occasion some difficulty and doubt, and may 
then constitute one of the more troublesome features in the interpretation of large 
series of scales. But the proportion of doubtful cases is very small and such can be 
eliminated from the series without danger of affecting disastrously the results. 
McMurrich’s plate vi, figure 2, represents a grilse in its third year. Here the 
nuclear area of the scale is abnormal and does not give satisfactorily the history of the 
first year. Other scales from the same fish would have given this in all detail. But 
we have to do apparently with a fish of stream type, which spent its third winter (marked 
second winter) in the sea, and was therefore toward the close of its third year. Its 
length (approximately 20 inches) is that which we have found uniformly characteristic 
of third-year grilse of stream type. 
SILVER SALMON, OR COHO (Oncorhynchus kisutch). 
[PI. xi; fig. 21, pi. XU.] 
The coho agrees with the sockeye and king salmon in having a dual habit during its 
first year. Certain of the young migrate to sea as soon as free-swimming, others, in 
unknown proportion, remain in the stream until their second spring. Fingerlings are 
present in all streams visited by this species throughout the summer, fall, and winter 
of the first year. If a seaward migration occurs in the fall, it has so far not been dem- 
onstrated. In the latitude of San Francisco yearlings are very numerous in all the 
smaller streams as late as March and April, and are often caught by trout fishermen 
during the early spring months. They remain in evidence several weeks after the 
appearance of the fry of the year, and may then be 3 to 4>^ inches long, being of the same 
size and general appearance as yearlings artificially reared in aquaria. Rather suddenly, 
on some spring freshet, they disappear from the stream. Some in their downward 
migration are often left stranded in overflow pools along the lower course of the stream. 
