I found a difference of from three and a half to six inches—­the smallest having the same silvery coat as the largest.  We cannot at all wonder at this difference, as it is a fact that the spawn even of the same fish exhibits a disparity in its fry as soon as hatched, which continues in all the after stages.  Although the throng of our smolts descend in April and May, we have smolts descending in March, and as late in the season as August, which lapse of time agrees with the continuance of our spawning season.  But in all these months we have an equal proportion (that is, a corresponding mixture) of large and small smolts.  I have earnestly searched for smolts in the winter months, year after year, and I can only say that I have never seen one, although I have certainly tried every possible means to find them.  I have seen fish spawning through the course of six months, and I have seen smolts descending through the same length of time.  Our return of grilses, too, exactly corresponds with this statement.  Thus a few descending March smolts give a few ascending May grilses; while our April and May swarms of smolts yield our hordes of grilse in June and July.  After July, grilses decrease in numbers till October, in proportion to the falling off of smolts from May to August.  At least these are my observations in our northern streams.”  They are observations of great value, and it is only by gathering together similar collections of facts from various quarters, that we can ultimately attain to a clear and comprehensive knowledge of the whole subject.

We gather from our most recent correspondence with Mr Shaw, (Letter of 8th June 1843,) that he does not regard the range in the spawning period to be followed by a corresponding range in the departure of smolts towards the sea, and in their return from it as grilse.  He has found a considerable diversity of time in the assumption of the silvery coating even among individuals of the very same family.  “I do not,” he observes, “recollect an instance where there were not individuals of each brood reared in my ponds, which assumed the migratory coating several weeks before the brood in general had done so; and these individuals would have migrated accordingly, and reappeared as grilse all the sooner.”  As the hatching and growth of salmon smolts and other fish, is regulated in a great measure by the temperature of the water in which they dwell, it is very probable that ova deposited late in the season, (say the month of March,) may, in consequence of the great increase of temperature, be hatched much more rapidly than those spawned in mid-winter, and so, by the end of a couple of years, no great difference will exist between them.  We remember that, in one of Mr Shaw’s earlier experiments, it is stated that he took occasion to convey a few ova in a tumbler within doors, where the temperature ranged from 45 deg. to 47 deg..  They were hatched in thirty-six hours, while such as were left in the stream of the pond, in a temperature of 41 deg., did not hatch until the termination of seven subsequent days.  The whole had been previously one hundred and six days in the water, under a considerably lower temperature.