SMOLTS 15 
in a week or ten days. In small streams the time 
may be shorter. 
When a salmon is stripped for hatchery purposes 
as many eggs as possible are pressed from the 
abdomen, and the fish is left in a collapsed and 
wrinkled state quite foreign to the naturally 
spawned-out kelt. At the hatcheries of the Pacific 
coast rivers in America each female is first killed 
and then slit open. A considerable proportion of 
the eggs so taken are not naturally ready for 
extrusion, and yet the percentage of loss through 
unfertilisation or subsequent death is small. In 
the natural process all extruded eggs are per- 
fectly ripe, and it is extremely unlikely that the 
percentage of unfertilised eggs is greater. Over 
against this, however, we have to reckon with the 
attendance of the hungry trout as the redd is being 
covered, and the subsequent attacks of creatures 
capable of penetrating amongst the gravel. With 
the conditions which obtain in the great majority of 
our Scottish rivers there is abundant evidence to 
show that the natural spawning of the salmon pro- 
duces a return in smolts sufficient to provide for a 
full stock. 
The egg, then, lodged in a dark crevice between 
stones of the river bottom, develops till the time of 
hatching arrives. The little fish, with its attached 
yolk-sac, its food supply, then wriggles out, and for 
a period of about fifty days continues to grow in 
those dark recesses. Towards the end of this period, 
as the yolk-sac has become much reduced, feeding 
through the mouth is commenced, and efforts for 
