27 
in fresh-water streams that one could cross the stream 
on their backs if he could keep his balance, ‘is true. 
The writer has seen salmon like that scores of times. 
It happens indeed every year in all good salmon streams 
where the primeval abundance of the fish has not been 
reduced by human agency. Now, imagine all these 
streams filled, as they are every spawning season, for 
weeks and months, with thousands, yes, hundreds of 
thousands of salmon, all crowded together where there 
is no food for them, and suppose at the same time that 
it was necessary to their existence to have food. What 
would be the result? The result would inevitably be 
that they would soon be driven wild with hunger, and 
would doubtless in their desperate extremity endeavor, 
if possible, to return to the ocean. Later on, if they 
continued to stay in fresh water, they would die of 
starvation, before the days of their spawning were 
accomplished, and ultimately the race would in conse- 
quence become extinct. 
It is evident therefore that the salmon must be en- 
abled to live without food in their fresh-water spawning 
streams. Otherwise it would be simply suicidal for 
them to go up the streams to spawn. In other words, 
if the salmon must of necessity go where there is no food, 
they must also of necessity be provided with the power 
of living without food. 
The common objection to the statement that salmon 
do not feed in fresh water, is one that comes up at once 
in every mind instinctively, namely, that it is not natural 
for an active creature like a salmon to go without food 
for so long a period as the salmon have to. The argu- 
ment, put concisely, is that it is unnatural that they 
should live so long as they do in fresh water without 
feeding, and consequently it cannot be that they do not 
feed there. The reply is that it is much more unnatural 
that, being compelled to feed in order to sustain life, 
the salmon should be sent into places to stay for months 
where there is no food to be had. This would be un- 
