73 
TAGGING SALMON IN ALASKA, 1923 
The record is important in that it indicates the possibility of a mingling, on the 
feeding grounds, of salmon from both the Asiatic and American shores of the north 
Pacific. 
The data indicate that there are certain differences in the rate of travel of dog 
salmon when compared with red salmon. A comparison of the results of the ex- 
periment with dog salmon begun at Unga Island, July 6, with the results obtained 
with red sa lm on tagged the same day, shows that the two species required almost 
the same time to make the trip to Bristol Bay — about 16 days. In the case of the 
fish taken in local fishing areas along the southern shore of the peninsula, however, 
the dog salmon took a noticeably longer time,' especially for the migration to Ikatan 
and Morzhovoi Bays. This may possibly be explained on the ground that most 
of the red salmon taken locally were actually bound for Bristol Bay and were 
therefore traveling more rapidly. 
HOMING INSTINCT IN TAGGED RED SALMON 
Sufficient evidence had been obtained previous to this tagging experiment to 
demonstrate the practical unanimity with which salmon return to their native 
stream at spawning time. The most direct evidence has been derived from numer- 
ous marking experiments performed on fingerling salmon before they had left 
fresh water. The recapture of these, when they have returned to spawn on reach- 
ing maturity, has been invariably in the stream in which the marking experiment 
was conducted. Such cases number many hundreds and comprise a body of 
evidence in itself conclusive. 
A less direct but equally important line of evidence is found in the fact that 
the salmon of different streams often possess obvious distinguishing characteristics, 
which mark them off as distinct races. The individuals of one race may be much 
smaller than those of another even when they are of the same age, or they may be 
longer and more slender; one may be consistently darker in coloration even when 
in the ocean, or the flesh may be redder and have more oil. Such differences could 
have arisen only by virtue of such a high degree of isolation and continuous in- 
breeding as would permit and foster divergence. 
Am ong the other differences which thus arise are distinctive growth habits, 
which may register themselves in the structure of the scales and can be recognized 
when these are subjected to miscroscopic examination. These growth habits 
include amount and rapidity of growth both in stream and sea, the length of resi- 
dence in fresh water, the final age at maturity, and many others. 
As our experiments have shown, red salmon that were schooling in the vicinity 
of the Shumagin Islands and in Morzhovoi and Ikatan Bays preparatory to then- 
spawning migration found their way when tagged to a large number of different 
rivers, some of which were located along the southern shores of the Alaska Penin- 
sula, others along its northern shores in Bering Sea. That the majority of them 
were bound for Bristol Bay was but a corollary from the fact that Bristol Bay 
produces by far the most extensive red-salmon runs of these northern waters. 
Tagged salmon were captured with the spawning runs of the Ugashik, Egegik, 
Naknek, Kvichak, and Nushagak Rivers, comprising all the red-salmon streams of 
Bristol Bay, and they were taken in sufficient numbers to warrant an inquiry as to 
