12 THE LIFE OF THE SALMON 
months’ time. The adult fish are no doubt in great 
numbers in those west coast American and British 
Columbian rivers, yet the rivers themselves are of 
vast size, and the amount of food for fry must be 
practically unlimited. In any case we have no evi- 
dence that smolts would starve in any of our rivers 
if they did not descend when they do. No doubt 
they get a greatly increased and varied amount of 
food when they do go into the sea, but the time 
when the migration takes place is the time when the 
best feeding season is commencing, and it seems to 
me necessary to take the natural instinct for a 
temporary marine sojourn into account as well as 
the need for increase of food. 
The time at which spawning takes place naturally 
influences the condition of the fry, and possibly, 
within limits, the time at which the smolt enters the 
sea. We have in Scotland at the present time, when 
account is taken of the Border rivers, considerable 
variation in the limits of the spawning season. So 
far as reports show respecting the times at which 
fish are noticed spawning in the various Scottish 
districts—which reports are published annually in 
Part II. of the Fishery Board’s Reports—there is a 
difference between the earliest and the latest mean 
periods of fully two months and a half. If, however, 
the Border rivers Tweed, Annan, and Nith are ex- 
cluded—and they have been subject to special and 
peculiar conditions—the difference in mean time 
between Scottish districts is reduced to one month. 
In other words, the height of the spawning season 
in the earliest river, viz., November 7, is one month 
