I think that its ultimate judgment will be a compromise.

Why should it be deemed improbable that a fish takes a fly from hunger at one time and from some emotional impulse at another ? To say that salmon flies are not like any insect is hardly a persuasive plea against the thought that they may seem good to eat. Only three or four generations ago men fished for trout, and caught them, with flies three or four times larger than the insects which, very clumsily, the lures were designed to imitate. Then, are salmon to be thought of as for certain differing from all other animals whose moods we understand in respect that they are incapable of irritation, or of curiosity, or of playfulness ? A remark by Mr. Rudyard Kipling, in a private letter to myself, may, I think, be mentioned without impropriety. Mr. Kipling speaks of " fish rising from nasty temper, which," he says, " I have seen a salmon do more than once." An implication obviously is that at other times they rise from another impulse. Indeed, there is reason for believing that sometimes they rise in a rage and eat at leisure what they seize. As the testimony of intelligent observers is not less valuable than deductions from theory, I will cite a striking case as presented in The County Gentleman. Narrating experiences on an Irish river, " Shannon " writes : " It is by ' spotting' fish first of all, and letting the prawn come down to them from a distance, that the best sport is obtained. It is very interesting to watch the habits of different fish. Some, though few, fly away at sight of the bait. The majority sulk, some even to the extent of allowing the prawn to touch them. Others go at it at once; but the most exciting time is with the fish who gets gradually angry. Fish of this sort begin by waving the fins and knocking the shrimp away, sometimes a yard or more; often they sail round and come back to the same ' lodge,' when the same thing is repeated; generally it ends in the fish being caught. Some fish are very savage. I saw one tear two prawns off the hook in succession, and eat them in mid-water. He was caught on a third prawn. There are different kinds of prawns; the pale pink ones are not well taken, but the short dark red ones, especially with eggs on them, are taken greedily.1'

Then, what are we to make of the familiar fact that on practically all the " autumn rivers11 throughout the United Kingdom sport invariably becomes brisk just before the close of the season ? As the final weeks are often a time of drought, the good sport cannot be regarded as proof of Dr. Barton's theory that salmon rise at fly only when they have been running and are fatigued. As I write, just after the close of season 1905, the rivers have been rather low for more than a month, and the fish have not been moving much; yet, as usual, the baskets of the final three weeks were the heaviest of the year. How is that to be explained away ? Does the irritability of the salmon wax as the time of spawning approaches ? It may; but one feels it more natural to believe that the activity of the fish is attributable to hunger following the gradual consumption of the stored-up sustenance of which Mr. Huxley speaks. Trout also rise particularly well just before the close of the season, and it is never supposed that in their case the impulse is other than that of sharpened appetite arising from increasing need of nourishment.

In all the writings of the men of science I have not found the slightest reference to a consideration that is of more importance than all the pros and cons we have surveyed. Although the origin of the salmon is not definitely known, it is generally believed that the fish has always been native to fresh water, and that at the early times of the species it lived in that water exclusively. If that be so, the salmon must originally have found all its food in lakes and streams. It feeds now, we know, in the sea, and its habits may have become modified; but is it unreasonable to believe that it may feed in the rivers and the lakes as well ? To believe this seems more than reasonable. It seems imperative. One of the most prominent doctrines of modern natural science is, that it is as difficult to be quit of a hereditary instinct as it is to acquire a character novel to the race. Is it not, then, necessary to assume that when a salmon rises at a fly, or seizes a sunken bait, he is, occasionally at least, repeating the action of remote progenitors that fed and flourished in the very pools where we are seeking sport?

The appetite of his eye may be greater than that of his palate; it may be a disembodied impulse, illusory ; perhaps it arises in racial reminiscence rather than in actual need. Still, appetite of some kind, true or false, it does really seem to be. Mr. Huxley, who viewed the salmon broadly, and not in their habits as they live for sportsmen, probably did not know that as a rule they rise particularly well at and about sundown. The possibility that the fish have a regular hour for feeding, or trying to feed, is conceivable ; but the theory that they have a regular hour for being in a rage could be accepted only as part of a revelation that universal life is a grotesque comedy.

The possibility that the fish rise from curiosity or in playfulness is not so easily disposed of. Certain animals, such as rabbits in the evening, frolic at regular times, and it is conceivable that salmon may have a similar wont. On the other hand, trout, which late in spring and throughout the summer rise particularly well in the evening, rise then, it is known, to feed ; trout and salmon are kin ; and it is not easy to believe that while one set of fish are rising to feed the other set are rising in frolic only.

All the considerations weighed, it does seem approximately certain that, though there may not be good digestion to wait on appetite, the salmon, when he rises, usually means to eat.