58 
BULLETIN OF THE BUREAU OF FISHERIES 
In the first experiment, started June 30, over 44 per cent of the fish tagged 
were taken in Ikatan Bay. This is correlated with a correspondingly low per- 
centage of returns from points to the eastward (5.2 per cent). With this exception, 
comparatively few of the fish tagged here were reported taken in Ikatan Bay. 
This is additional evidence that of the fish found in Morzhovoi Bay those that 
are destined to enter local streams, both in Morzhovoi Bay and to the eastward, 
are found along the eastern shore of the bay and mainly during the latter part of 
the season. 
RATE OF TRAVEL 
In making an analysis of the rate of travel it has become apparent that certain 
facts affect the reliability of the returns from Bristol Bay. These will therefore 
be discussed after the returns from other localities have been examined. The 
data will be found in Tables 30 to 35. 
It is obvious that the rates of travel, as determined from the elapsed time 
between tagging and the reported recapture, will be only a minimum rate. In the 
first place, the assumption that the fish travel in the most direct line is not reason- 
able. Then, too, there will necessarily be some delay between the time that the 
fish is first taken in the trap and the time that it is found and the recapture re- 
ported. Traps are seldom lifted more than once a day (frequently a much longer 
interval elapses), and since the fish may enter the trap at any time during the 24 
hours or more between one lift and the next, it follows that the elapsed time between 
liberation and reported recapture will be, on the average, at least 12 hours longer 
than was actually required by the fish to make the journey. It is certain, there- 
fore, that the actual rate of travel of these fish is somewhat higher than our figures 
show, but just how much higher it is impossible to say. 
FROM THE SHUMAGIN ISLANDS 
The important migration from Bnga Island to Ikatan and Morzhovoi Bays, 
composed, as we have seen, largely of fish bound to Bristol Bay, requires approxi- 
mately 7 days. After each tagging at Unga Island fish were reported from Ikatan 
and Morzhovoi Bays, or both, within 4 or 5 days, occasionally in 3 days, and were- 
reported thereafter for about 2 weeks. Tagged fish were taken in the greatest 
numbers within 5 or 6 days after the tagging, the numbers diminishing gradually 
until only an occasional tag was reported later than 2 weeks after the experiment 
was begun. 
An inspection of some of the more typical tables will show clearly the extremely 
skewed type of frequency distribution given by these data. Not infrequently 
there were more recaptures reported on the first or second day than on any subse- 
quent date. (See returns from Ikatan in Tables 3 and 6.) No conspicuous differ- 
ence is noticeable in the time required for the fish to go from Unga Island to Ikatan 
or to Morzhovoi Bay, nor is there any indication that the rate of travel varies in 
any regular manner during the season. The weighted means of the median times 
found in the tables are, from Unga Island to Ikatan Bay, 6.6 days, and to Mor- 
zhovoi Bay, 6.9 days. The distances are approximately equal — about 120 miles — 
so that if we assume that the fish traveled by the shortest and most direct route 
their rate of travel was not less than about 18 miles per day. 
