KEOKUK DAM 
99 
at the beginning and end of seasons) during this year and a half varied from 2 to 89; 
but there were only two weeks with more than 58 lockages. During the first half of 
1916 the largest number in one week was 49. 
Taking the average of lockages per day as 5 and the average time per lockage 
as 5 minutes, there was an opening for fish of 25 minutes a day, on the average, or 
2/ hours in a week. The maximum opening in any week during the year and a half 
in question, excluding two extreme cases, was about 4 hours. Or, to present the 
figures in one more form and using the same average time per lockage, there was, 
during the year 1915, an opening for fish for a period of time equivlent to 5 % days, 
and during the first half of the calendar year 1916 (including July 1) lji days. 
EXTENT OF FISH MOVEMENTS DURING INTERVALS OF FREE PASSAGE 
The facts already adduced show that there was but little time when a fish had 
even the slightest chance to pass out of the upper end of the lock. It has also been 
indicated that the situation was unfavorable during these brief intervals because of 
the lack of current and the relative smallness of the exit. Certainly, the lock is 
not entitled to be considered a fishway unless, in spite of these adverse circumstances, 
there is a dense mass movement of fishes leaving the lock when the upper gate is 
submerged. The evidence available as to the extent to which fishes move over this 
upper gate will now be considered. 
It will be understood that a trammel net consists of three nets put together, 
sandwichlike, the middle one, called the “web,” being of small mesh and the two 
on the outside, called the “walls,” being of mesh much too large to stop any fish that 
it is expected to catch. A fish striking the net with appreciable force pushes the 
web through the opposite wall and thus pockets itself. Figure 7 will perhaps show 
this more clearly than the description. 
A trammel net with web of 2-inch (5-centimeter) mesh between knots and wall 
of 8-inch (20-centimeter) mesh was supported on the gate by means of three 13-foot 
(4-meter) pieces of gas pipe, these having the net tied to them and being in turn sup- 
ported in three holes bored into the respective ends and the middle of the gate. 
Wedges held the bottom or lead line of the net fast to the gate, and corks kept the 
other edges at the surface of the water. During part of the first month the cork line 
would go a little beneath the surface at one end, but this was corrected by doubling 
the number of corks. When the gate was down the only possible passage was closed 
by the net, except as hereinafter stated. (Fig. 8.) Operations of the gate were made 
especially for this work. To give fish an opportunity to enter the lock the lower 
gates were left open a half hour or more (except on April 3) before closing them and 
filling the lock. Sometimes they were open 8 or 10 hours; the average was about 
2 hours. Leaving the lower gates open for this length of time made the circumstances 
slightly more favorable than is normally the case; but it would be entirely possible 
to have the lower gates left open nearly all the time, and the circumstances during 
the experimental work were, therefore, no more favorable than they might readily 
be made were it found worth while to do so. 
A fish might pass the net at either end through a space between the net and the 
wall of the lock, possibly 1 % feet (% meter) wide at the widest part. A trammel net 
is designed to catch fish on their first contact with the net, and there is probably 
little passing of the fish along the net toward the ends; however, a few individuals 
undoubtedly went through at those places. The only other way that a fish too large 
