For this reason, at the opening of the season anglers seek their sport along the northern shores. It is a true instinct that guides them thither. At the opening of the season trout are most frequent where the water is least cold. Why, then, it will be asked, does sport along the shore sometimes fall off when spring has advanced a stage ?

The answer will arise on consideration of a characteristic in which water is in harmony with all substances other than explosives such as gunpowder and nitroglycerine and dynamite. Heat does not act upon it so quickly as it acts upon earth, or so slowly as it acts upon wood ; but water is like all of the substances which we meet in the fields or in the woodlands in that its permeation by heat or by cold is gradual. Small bodies of water, which are more quickly heated than large bodies, are also more quickly chilled. By the middle of April the general body of water in a lake is of a sensibly higher temperature than it was in January; but by that time another change, a change affecting the habits of the trout, has come over the conditions of the lake. Even as the shallows were the first places to be warmed when the sun waxed early in the year, they are the first to become cold when the frosts, as they are wont to do, return. It is always on the shallows that the first ice appears. Thus, a " cold snap " in early spring will cause the temperature of the shallows to fall below the mean temperature of the lake. A similar result comes of another cause, which seems to have escaped general notice. During a succession of sunshines early in the year, the shallows near the shore have day by day been made warmer than the deeps; but all the time that this has been going on it is the deeps that have been most surely gaining. Nightly the shallows have lost most, if not all, the warmth of the day; but by night as well as by day the deeps have been storing nearly all that they received. They have been retaining all except the comparatively small portion which the laws of nature called upon them to give up during the nights ; while night by night, at the same instance, the shallows have had to part with nearly all the warmth which sank into them during the day. Consequently, although the shallows continue to be warmed day by day, there comes a time when even in the full shine of the sun at noon they are chiller than the deeps.

That time begins about the middle of April. It is then that sport along the shore falls off. The trout have neither ceased to feed nor become more wary. They have simply sought more comfortable quarters in the deeps.

It goes against the grain to be frequently referring to one's own experience, and in this book I strive to keep such references as few as may be, making the narration, as a rule, oblique; but sometimes the bearing of personal witness is inevitable. It seems to be so now. The theory about the change in the haunts of trout which has just been set forth is derived from daily observation from the opening of a season. Well-filled baskets were the rule all through March and the first half of April, and these were the fruits of fishing either from the banks or from a boat drifting along the banks ; but suddenly this good fortune was at an end. It became as difficult to catch a brace of trout as it had been to catch a score. One morning when, there being no wind, the lake was placid, I noticed that, while a strip of water extending outwards at least thirty yards from the shore was undisturbed by rises, beyond that trout were moving everywhere. Seeing that the fly on the water was a small insect with grayish-white wings and a black body, I put on a cast of midgets, rowed out into the middle of the loch, and had very good sport indeed. The spell was broken. The manner of the breach was rather surprising to myself as well as to the hospitable household with whom I was staying on a holiday. There, as throughout Scotland and England generally, it is a traditional belief that, excepting on lakes which are shallow all over, trout away from the shores cannot be induced to rise at a fly. I had accepted the tradition; but the results of the experiment that calm morning were, in a pleasant sense, disturbing. It had struck me that, as trout in the middle of the loch were rising at real flies, there was no reason for thinking that they would fight shy of artificial ones; and the expectation had been justified.

Still, the experiment was not yet complete. Thinking that my host and hostess, when I spread out before them the produce of the deeps, which were believed to have no produce at all, would say, " O! but you would have caught them along the shore too, if you had fished there : all that has happened is that the trout have come on the rise again," I tried the shore, tried it quite fairly for half an hour, and did not get a single rise.

Out upon the deeps I pulled the boat once more; and the basket was two or three pounds heavier when it was time to return.

My theory about the influence of the temperatures seemed demonstrated to the full; but would it always hold ? Day after day for many days I put it to the test, and the results were not easily interpreted. On the deeps when the weather was calm a few trout were usually caught, and often when there was a wind light enough to cause only a gentle ripple I had a good many more than a few; but when the wind was high enough to make the wavelets break into spray there was practically no sport at all. On these days, too, the trout began to come whenever, in its drift, the boat neared the shore. For a time this was disconcerting ; but it was not, I think, inexplicable. Spring had been moving on; even at night, the weather had been mild, and usually in daytime warm; in the course of their changes, the temperatures of the water in its various parts had either been approximately equalised, or had been raised so much that none of them was too rigorous for the comfort of the trout. Other things being equal, it is the shallows that the fish prefer. We see this on rivers. Trout are to be found in canals deep enough for large ships; but that is because, being there, they cannot help themselves. In a river their preferences are unmistakable. It is not in the deep channels that the wise man seeks them. A considerable pool in the middle of a river they will not shun, being able, from that ambush, to see all the living dainties that come towards them over the rim of gravel; but in the very deep channels they are absentees or merely "passing through." It is undoubtedly the shallows that they prefer: the tail of the rapids, the rapids themselves, and, in the slowly-moving stretch which is usually bounded by the dykes of a miller's dam, the sides, where the mud-banks shelve upwards among the sedges. Their habits in a lake are similar. They tend towards the deeps when these are the least uncomfortable parts of the water; but they prefer the shallows at other times.