just seen that smolts return to the rivers as grilse, (of the weights above noted,) during the summer and autumn of the same season in which they had descended for the first time to the sea.  Such as seek the rivers in the earlier part of summer are of small size, because they have sojourned for but a short time in the sea:—­such as abide in the sea till autumn, attain of course a larger size.  But it appears to be an established, though till now an unknown fact, that with the exception of the early state of parr, in which the growth has been shown to be extremely slow, salmon actually never do grow in fresh water at all, either as grilse or in the adult state.  All their growth in these two most important later stages, takes place during their sojourn in the sea.  “Not only,” says Mr Young, “is this the case, but I have also ascertained that they actually decrease in dimensions after entering the river, and that the higher they ascend the more they deteriorate both in weight and quality.  In corroboration of this I may refer to the extensive fisheries of the Duke of Sutherland, where the fish of each station of the same river are kept distinct from those of another station, and where we have had ample proof that salmon habitually decrease in weight in proportion to their time and distance from the sea."[19]

    [18] These two specimens are now preserved in the Museum of the
    Royal Society of Edinburgh.

[19] The existence in the rivers during spring, of grilse which have spawned, and which weigh only three or four pounds, is itself a conclusive proof of this retardation of growth in fresh water.  These fish had run, as anglers say—­that is, had entered the rivers about midsummer of the preceding year—­and yet had made no progress.  Had they remained in the sea till autumn, their size on entering the fresh waters would have been much greater; or had they spawned early in winter, and descended speedily to the sea, they might have returned again to the river in spring as small salmon, while their more sluggish brethren of the same age were still in the streams under the form of grilse.  All their growth, then, seems to take place during their sojourn in the sea, usually from eight to twelve weeks.  The length of time spent in the salt waters, by grilse and salmon which have spawned, corresponds nearly to the time during which smolts remain in these waters; the former two returning as clean salmon, the last-named making their first appearance in our rivers as grilse.

Mr Young commenced marking grilses, with a view to ascertain that they became salmon, as far back as 1837, and has continued to do so ever since, though never two seasons with the same mark.  We shall here record only the results of the two preceding years.  In the spring of 1841, he marked a number of spawned grilse soon after the conclusion of the spawning period.  Taking his “net and coble,” he fished the river for the special purpose, and all