68 
BULLETIN OF THE BUREAU OF FISHERIES 
day was 13. Under these circumstances it is to be expected that the variation in 
the median time required to make the trip would be much greater in 1922 than in 
1923. In spite of this greater variability it is quite apparent that in general the 
results obtained in the two years are fairly similar. Experiments started on approxi- 
mately the same dates in the two years gave results so similar as to leave no doubt 
that the conclusions drawn apply to the migration each year. 
What determines the rate at which salmon travel in the ocean during their 
migration from the feeding grounds to the spawning grounds ? It has been shown 
repeatedly that the rate of travel is distinctly more rapid in the case of those fish 
having a longer journey to make. The data on the migration to Bristol Bay has 
shown, further, that with the same distance to go the fish travel more rapidly with 
the approach of the spawning season. It seems possible to conclude, therefore, 
that the rate of travel is the resultant of two main forces — the relative maturity of 
the sex products and the distance from the spawning grounds. Other factors may 
possibly affect the rate of travel, but of these we know nothing at present, and it 
seems more than likely that the two mentioned are of primary importance. 
PERCENTAGE OF RETURNS 
The percentage of returns serves, although very imperfectly, as a measure of 
the intensity of fishing. If, out of 1,000 fish tagged and liberated in a given locality 
on a certain date, 500 are subsequently recaptured, it is fair to assume that approxi- 
mately 50 per cent of the untagged fish found in that locality at that time were also 
taken. It must not be forgotten, however, that this is, at best, a minimum figure; 
that various factors are at work which keep the percentage of returns down and 
which, so far as we can see, will never raise it above the percentage captured of 
untagged fish with which they were associated at the time of liberation. These 
factors have been mentioned before and include the possible mortality of fish that 
have undergone the experience of being tagged, the retention of tags by fishermen 
and others as souvenirs, the loss of tags as a result of being detached between the 
time of tagging and recapture, and the failure on the part of the men engaged in 
handling the fish to notice the tags. In so far as our understanding of the in- 
tensity with which the important commercial runs of salmon are fished is concerned, 
these figures are also affected and reduced by the fact that some of the fish go to 
small streams in which no commercial fishery is maintained. 
In spite of the fact that the figures can not be considered fully reliable as a 
measure of the intensity of fishing, certain very interesting facts are apparent from 
the data given in Table 41. 
