108 THE LIFE OF THE SALMON 
and more plentifully in certain places and under 
certain conditions than others. It is natural, there- 
fore, to suppose that the kelts which remain longest 
in a river after spawning are, in their sickly and 
often wounded state, more likely to become infected 
than those fish which quickly leave fresh water for 
the sea. The fungus spreads quickly when fish 
begin to die, since saprolegnia flourishes much better 
on dead than on living tissue. It is noticeable that 
diseased fish are very commonly of large size, and 
that salmon disease is most conspicuous in large 
rivers. It is in large rivers that large kelts remain 
longest after spawning, and there is considerable 
evidence to show that these late kelts are fish which 
ascended fresh water in summer and autumn. Hence 
I would offer the observation that loss of fish from 
salmon disease is less likely to injure the stock of a 
river when the river maintains a good run of spring 
fish. In the immense rivers of British Columbia, 
Alaska, and California the kelts of salmon which have 
ascended many hundreds of miles are reported to die 
without exception; only those fish which do not 
ascend to the upper waters seem to be able to return 
to the sea. The great runs of fish in those Pacific 
coast rivers do not occur till late summer, but the 
expenditure of energy necessary for so great an 
ascent must produce a much greater drain upon the 
systems of the fish than is found in any upper water 
fish in our country, and in all probability exceeds 
the lost of nutriment sustained by our spring fish 
which remain long in fresh water. But for the 
excessive fishing of salmon traps in the Pacific coast 
