230 BRITISH FRESH-WATER FISHES 
enough ground upon which to base the theory that salmon 
devour salmon. It is quite true that certain fish, such as pike, 
trout, and others very low in the scale of vertebrate animals, 
are at times pressed so hard by hunger as to transgress the 
law which the fiercest mammals respect, and the stronger 
individuals make prey of the weaker of their own kind ; but it 
is tolerably certain that, so soon as a kelt feels the seasonal 
return of appetite, he makes off to the sea to satisfy it. The 
only circumstance which might make him turn cannibal would 
arise in the event of his being imprisoned in the river by 
drought, in which case no doubt the kelt would not be 
particular about his diet; he would stay his stomach with 
whatever edible objects he could find. 
Nobody, I suppose, has ever accused fresh-run salmon of 
cannibal propensities, and anglers often receive the impression 
that kelts are far more ravenous than fish which have just 
entered the river from the sea. The fact is that, in the spring 
months, when kelts are descending to the sea, and clean fish are 
beginning to run from it, probably more kelts will be landed 
by fly or spinning bait than clean fish; the reason being, not 
that the kelts are more hungry, but that they are far more 
numerous than the clean fish. 
My own belief, founded on prolonged observation, is that 
the clean-run fish takes the bait far more readily than the spent 
fish, not because he is the hungrier of the two, but because his 
vitality is higher and his predaceous instinct more alert. 
There are good grounds for believing that, from the time the 
salmon leaves the sea until after the operation of spawning, it 
takes no nourishment. Instances may occur of salmon in fresh 
water swallowing edible objects, but the fact remains that such 
objects have never been detected in their stomachs, nor the 
débris of them in their intestines. That salmon in rivers and 
lakes seize moving objects, which may or may not be edible, 
such as minnows, worms, and artificial flies, has been cited as 
proof that they feed in fresh water, and the uniform emptiness 
