FISHERIES OF ALASKA IN 1909. 
35 
The first visit to the lake was made June 7, when from an examina- 
tion near its foot it appeared to be clear of ice, whereas on May 31, 
1908, it was still almost entirely covered. The work of installing 
the rack was begun June 17 and the temporary crew was taken away 
on the 22d. The rack was not made entirely tight until the 2d of 
July. The few salmon first appearing, however, show no tendency 
to seek the apertures under the chain at the bottom of the web, but 
find their way readily through the gates. The same or greater trouble 
was encountered with the floating drift as in the previous season. 
This drift is composed of dead leaves, grass, twigs, and small limbs 
of trees set afloat from the shore by the rise of the lake and brought 
down by the current and favorable winds. It arrived at the salmon 
rack in greatest abundance between June 26 and July 2, and during 
this period strenuous exertions on the part of the crew were neces- 
sary to keep the web sufficiently clear to relieve the pressure against 
the piles. As it was, several of these were bent over and had to be 
held by guy lines. The drift gradually lessened in amount and 
finally disappeared almost entirely. 
The first tally was made June 28 and the last on August 1, the gear 
being taken out the next day. In 1908, with a large run and the 
necessity of developing methods of making a proper and at the same 
time rapid approximation of the numbers, it was thought necessary 
to estimate from the frequent and regular tally of a uniformly flow- 
ing stream of fish. During the present season no estimates of any 
sort were made. The much smaller run and the fact that three gates 
were available made an exact count a comparatively easy task. To a 
certain degree this year’s figures corroborate the count of the pre- 
ceding season, for the daily quotas during the height of the run are 
consistent with each other during the two seasons, and it was obvious 
by mere inspection of the daily abundance about the gates that the 
1909 run did not approach that of 1908. It is probably possible to 
make an actual tally in any season, even of the largest runs. The 
salmon may be held below the rack without injury for a time, and 
by increasing the number of tally gates the fish may be delivered 
more rapidly. It is merely a question of a sufficiently strong and 
tight barricade, with plenty of gates and tallymen. 
An interesting coincidence is seen in the fact that the highest daily 
tally, the crisis or height of the run, occurred on the same day of the 
month, July 14, in each of the two seasons. While this height of the 
run will probably be found to occur usually near the middle of July 
at the lake with as much regularity as the entrance of the run into 
the bay, which usually occurs about the first of July, it is not to be 
expected to recur continually on any given date. It is suggestive, 
