76 THE LIFE OF THE SALMON 
matter and the wide circulation of notices respecting 
marking, and the offer of rewards, I do not think 
it likely that many records are lost in this way. 
Further, the examination of salmon scales seems to 
indicate that some fish may pass over two or it may 
be three consecutive spawning seasons. 
So far as the Scottish returns go, it is clear that 
the majority of fish spend much more time in the 
sea than in fresh water; and the Irish returns, 
although fish showing the short period are in the 
majority, and their average absence about six months, 
nevertheless yield a proportion of long period fish, 
so that here also most time is spent in the sea. It 
is quite conceivable, therefore, that with our netting 
in the mouths of rivers and close in shore only on 
the coast at a definite season we fail to trace a 
proportion of our marked fish. Yet in contradis- 
tinction from this it has happened that from some 
36 kelts caught by rod and marked in the Upper Tay 
alone, as many as 22 per cent. have been recaptured. 
The stock of salmon is, however, entirely depen- 
dent upon the fish which enter the rivers, since 
salmon do not spawn in the sea and the ova are incap- 
able of development if shed in the sea water. Hence 
the importance of moderation in our netting of rivers 
and estuaries of rivers. It is, of course, clear that 
amongst netted rivers those of large volume with 
wide estuaries are more capable of maintaining a 
suitable stock of breeding fish than small rivers with 
narrow mouths, since in the former a proportion of 
each run of fish is more likely to pass the nets, 
provided the gauntlet of netting is not too long to’ 
