In late autumn or in winter the salmon spawn in the rivers. The female
makes a shallow trough in the gravel by moving her tail from side to
side, and therein lays many eggs. The male, who is in attendance,
fertilises these with the milt, and then the female covers them deeply
with gravel. The process is repeated over and over again for a week or
more till all the eggs are shed. For three to four months the eggs
develop, and eventually there emerge the larvæ or _alevins_, which lurk
among the pebbles. They cannot swim much, for they are encumbered by a
big legacy of yolk. In a few weeks, perhaps eight, the protruding bag of
yolk has disappeared and the _fry_, about an inch long, begin to move
about more actively and to fend for themselves. By the end of the year
they have grown to be rather trout-like _parr_, about four inches long.
In two years these are double that length. Usually in the second year,
but it may be earlier or later, the parr become silvery _smolts_, which
go out to sea, usually about the month of May. They feed on young
herring and the like and grow large and strong. When they are about
three and a half years old they come up the rivers as _grilse_ and may
spawn. Or they may pass through the whole grilse stage in the sea and
come up the rivers with all the characters of the full-grown fish. In
many cases the salmon spawn only once, and some (they are called _kelts_
after spawning) are so much exhausted by starting a new generation that
they die or fall a victim to otters and other enemies. In the case of
the salmon of the North Pacific (in the genus _Oncorhynchus_, not
_Salmo_) all the individuals die after spawning, none being able to
return to the sea. It must be remembered that full-grown salmon do not
as a rule feed in fresh water, though they may be unable to resist
snapping at the angler's strange creations. A very interesting fact is
that the salmon keeps as it were a diary of its movements, which vary a
good deal in different rivers. This diary is written in the scales, and
a careful reading of the concentric lines on the scales shows the age of
the fish, and when it went out to sea, and whether it has spawned or
not, and more besides.