146 
BULLETIN OF THE BUREAU OF FISHERIES. 
commercial fisheries would indicate a maximal speed of three or four times that given by 
my highest rates. It is very probable, therefore, that the lack of uniformity in speed 
shown in the table is due to days consumed in ways not accounted for by the direct run 
through fresh water in the course up the river. 
An interesting side light is thrown on these observations if the speed for all is 
computed on the basis of the average speed made by number 76, the highest on the list. 0 
Table vi presents the results of this recomputation. 
Table VI. — Results of Computing Time Actually Taken in Run, on Basis of Average 
Speed of 7.5 Miles a Day. 
Species and number. 
Distance 
traveled 
in miles 
from 
point of 
liberation. 
Days out. 
Days required 
to cover 
distance at 
an average 
speed of 7.5 
miles a day. 
Days un- 
accounted 
for. 
Silver, 75 
210 
29 
28 
1 
Silver, 89 
210 
30 
28 
2 
Silver, 79 
210 
33 
28 
5 
Silver, 97 
210 
33 
28 
5 
Steelhead, 124 
210 
33 
28 
5 
Chinook, 80 
15 
1 1 
2 
9 
Steelhead, 98 
210 
S2 
28 
24 
Steelhead, 125 
70 
35 
9 
26 
Chinook, 123 
15 
±31 
2 
29 
Silver, 87 
70 
57 
9 
48 
Chinook, 113 
0 4 
6 
0 
t > 6 
o Downstream. b Had not yet left tide water. 
I fully recognize that table vi is based on an assumption. Nevertheless, it can 
not at present be displaced by observed facts, and serves better than an.y other method 
devised to illustrate the great discrepancy in the time consumed by numbers 80, 87, 98, 
123, and 125. The last column of the table shows that these particular fishes must have 
played around in the lower waters of the Columbia. Certain of them have not gone 
beyond tide water — for example, 80 and 123. This last fish has taken a whole month 
to go only 15 miles up the river. By the computation there are three others that have 
about the same time available for playing around or resting quietly somewhere, and the 
history of number 123 renders it quite probable that they all spent this extra time in 
tide water. 
We have, therefore, from this experiment two series of facts that throw light on 
the life history of salmon in tide water, namely, the etching or corrosion of the aluminum 
marking buttons and the probable time consumed by the salmon after they were marked 
at the state trap before they began the strictly fresh-water journey. Both observations 
show an unexpectedly long time in tide water, i. e., as long as 30 days (chinook number 
80) or even 48 days (silver number 87). 
Rutter 6 has advanced the theory that salmon make the journey through tide water 
by running up during the ebb and down during the flood tide, stemming the current each 
a It is evident from the slight corrosion of the button of this fish that it spent some time in brackish or salt water. 
It made, therefore, a really higher average speed during the time in fresh water. 
b Rutter, Cloudsley, op. cit., p. 122. 
