SMOLTS 11 
development of the species salmon as we now know it 
the bracing qualities of the sea, with its rich feeding, 
are absolutely necessary. I am aware that some are 
inclined to insist on the importance of the amount 
of feeding in our streams, and to go the length of 
suggesting that the stock of a salmon river should 
not be allowed to increase beyond the point where 
the smolt food fails to go round. To this I would 
reply that I would not limit the supply of adult fish 
on the chance of the parr starving, first because parr 
are evidently able to do a good deal of starving, and 
secondly because no man can well estimate the 
amount of parr food in a river. I should start to 
kill off the trout first. Also it seems to me that 
the amount of variation in the times and seasons 
when parr can go into sea water is probably an 
excellent provision against any danger of over- 
crowding. 
Herr Dahl, in Norway, considers that the parr 
there leave the rivers when one year old. I confess I 
am not satisfied that he is correct, and an examination 
of the scales_of Norwegian parr which I have had an 
opportunity of seeing—thanks to Mr. H. W. John- 
ston—supports this view; yet we know that the 
severe winter conditions of Norway modify very 
materially the habits of the adult fish. The fry of 
other salmonids seem to have different habits from 
those observed in Britain. Mr. Rutter has investi- 
gated the matter in the Sacramento river of the 
Pacific coast of America, and finds that there the 
fry begin to descend whenever they can swim, and 
commonly reach brackish water in about three 
