96 THE HABITS OF THE SALMON. 



successful issue of the spawning season two or 

 three years previously. Let us therefore inquire 

 by what cause the spawn and spawning grounds 

 are liable to be affected. All the spawning 

 grounds may have held their full complement of 

 spawning fish, and may have been well covered with 

 ova before hatching-time. But if heavy floods 

 appear, the eggs are washed away and destroyed. 

 Again, fish may have spawned in good places 

 which, from want of rain, have been left high and 

 dry; that also would be fatal. In a good spawn- 

 ing season, the river remains at a medium height 

 from the time the fish deposit their eggs to the 

 time of hatching. That there are good and bad 

 grilse years goes without saying, but we generally 

 anticipate a good grilse year two or three years 

 after a good spawning season. A good grilse 

 year is also dependent in a great measure on the 

 state of the water at the time they run. There 

 may be plenty of fish which cannot make head- 

 way for want of water, and so the greater pro- 



