590 
CHARLES WILSON GREENE 
Of the 18 steelhead that were marked and liberated 5 were 
re-taken. Number 116 was caught four miles down the river 
about four hours after it was liberated. The remaining steel- 
head were caught, one in the region of the point liberated, one 
seventy miles up the river, and two were re- taken two-hundred 
and ten miles up the river. Of the last two, one, number 124, 
was out thirty-three days and bore a button that showed only 
slight corrosion. The other, number 98, was out fifty-two days 
and its buttt)n showed marked corrosion. Evidently the former 
spent little time in brackish water, while the marking button of 
the latter indicates a long contact with sea-water. The steel- 
head number 125, taken seventy miles up the river after thirty- 
five days, shows a history of contact with salt water similar to 
that of number 98. 
THE AVERAGE SPEED OF MIGRATION 
This experiment was launched in tide water. Hence the speed 
of migration is influenced by the factor of acclimatization of the 
sea run salmon to fresh water. Undoubtedly a very much 
higher speed is attained in fresh water than is accomplished 
in making the journey through the tidal region. A number of 
instances have been given to show that the fishes spent much time 
in the brackish water after their marking. It is safe to assume 
that salmon travel at a fairly uniform speed in fresh water when 
different individuals are compared under similar conditions. In- 
spection of table 1 suffices to show that either the assumption is 
untenable or that a number of the salmon have not made direct 
runs up the river. Undoubtedly the latter represents the facts 
in the case, and the corrosion of the marking buttons is corrobora- 
tive proof. If the speed of the fishes is supposed to be uniform 
through the group and is computed on the basis of the average 
made by those representing the highest in the list, then, as table 1 
shows, there will be for each fish, a number of days unaccounted 
for. In some cases, numbers 98, 125, 123 and 87 the unaccounted 
for days amount to 24, 26, 29 and 48 days respectively. These 
are the days which represent the period of adjustment of the 
