looked no prettier out of the water than in it, but I pardoned his appearance and was much delighted with his capture. He was an old cock with a large hooked beak, and he ought to have weighed 13 lb. instead of 10. I had some of his scales examined, and the verdict was as follows; age 8 to 9 years, the first 3 of which he had spent in the river. Since then he had spawned two or three times. His removal seemed popular in the river, as his place was immediately taken by three others, overcrowding which Joe would never have allowed in his lifetime. These fish, however, moved up two days later.

Another special friend of mine was Gamlehaga Bill. He took one of my red quills quite early in our acquaintance, but the hook broke in his mouth. About a fortnight later I very nearly persuaded him to take a blue pill fished wet. I got no further rise out of Bill, and the last time I saw him was the Sunday morning on which I left the river. We gave him a parting present of a 'digestive' biscuit, floated in tempting pieces over his head. There was no hidden danger in the biscuit, but Bill thought there was, and at the second piece he turned sulkily away into the stream, and there we left him.

I was never able to discover any difference in the attitude of sea trout which were ready to take a fly from that of those which were not. They were always lying on the bottom and showing no movement except an occasional wave of their bodies from side to side. It was necessary to keep out of sight, though the natives had the common fault of being careless in this respect. Sea trout see a human being fairly quickly, and they have an annoying little trick of lying still and pretending to have seen nothing at all.

On several occasions I was able to watch every movement of the fish while the fly was floating over it. The trout in each case made no sign until the fly was passing over its head. If it meant to rise it turned quickly round on noticing the fly, and shot down stream and towards the surface, until it reached a point just below the fly. It then steadied itself in a vertical position before actually making the rise. The fly was taken from three to six feet below the fish's resting-place. The first movements were very rapid, but the actual rise was made quite slowly and in several different ways. The whole move

ment reminded me of a man in the deep field, racing back towards the boundary to make a catch from'a big hit, and steadying himself as he gets under the ball. Unfortunately, even sea trout sometimes misjudge and drop the catches. In making the rise the fish either just breaks the surface with its lips, or makes a head and tail rise, or rolls right out of the water like a porpoise, and takes the fly as it turns downwards again. The smaller ones now and then jump clean out of the water, drown the fly, and then pick it up from below. In some cases-apparently when he is too late to reach a point below the fly-the fish gulps it in with his head pointing down

stream.

The Norwegian gaffer told me that a sea trout which missed the fly did not rise again for ten days. How he made this calculation I do not know. At any rate it was not always correct. It was the exception to rise a fish a second time, but members of our party succeeded in doing so occasionally. Of all the good moments one has in this kind of fishing, I think the moment of the actual rise is the best. The sight of a great silvery trout rolling solemnly out at the floating fly sends down my back a fresh quiver of surprise, excitement and delight every time I see it. Great self-restraint is needed to prevent striking too soon. The actual rise is made so slowly that it is almost impossible to wait for what seems an age before the great mouth closes and the fish disappears. That is the moment to strike, and to strike hard. It is necessary to use gut strong enough to stand a considerable jerk. If all goes well and the hook is driven home at the right moment, I feel that rare sensation of having hit a half volley in the middle of the bat. The fish who merely shows his lips is the easiest to hook, because the rise is over so quickly that one has no time to strike too soon. On English chalk-streams it is rare to hook a brown trout far back in the mouth; but luckily sea trout appear to take the fly, if they take it at all, with a confident gulp which gives one the chance of driving the hook into the tough angle of the mouth, or perhaps (which is best of all) into the tongue.

It is difficult to understand why sea trout take the dry fly as readily as they do. On this river there was not often much natural fly on the water. On certain days a

fair sprinkling of grey duns, some red-bodied duns, and from time to time some small dark fly, appeared for an hour or two. Towards the end of July, at about 9.30 p.m., numbers of large sedges hatched on the banks, but these very seldom got on to the water during daylight; whether they did so late at night I do not know. I opened a number of fish, but the results, on the whole, were meagre. I found occasionally a good many grey dun nymphs and subimagos in their stomachs. In many there was nothing at all. In the majority there were a few small grey dun nymphs with the wings not much developed, corresponding in size to about 00 hooks. I also found in different fish one or two blue-bottles, a few sedges, a daddy-longlegs and some animals which looked like short, pale-green centipedes. I very seldom saw any number of sea trout feeding on natural flies; and it was unusual to see a large fish rise at all. If a rise was seen, the fish could generally be tempted by the artificial fly.

The best flies to use were Wickham's Fancy, Cinnamon Sedge and Grey Quill, size No. 10. It is advisable to take rather a larger size for the late evening, and rather a smaller size for days when the fish are coming short. The Grey Quill was an excellent fly, but it had one disadvantage; it was very hard to see in rough water. One fish, who was quite easy to watch, left his place and carefully examined a Silver Sedge, but thinking, I suppose, that it was too large for him at that time in the morning, he dropped back without rising. I gave him a smaller Grey Quill, and he took it as though he liked Grey Quills. The fight which followed was in rough water full of rocks and miniature waterfalls. The fish was fresh-run and very strong. He kept fighting for the pockets behind the rocks in which to get his wind, while I tried to guide and pull him out of the pockets into the little falls and runs, and so increase his exertions and prevent his getting any rest. To add to my difficulties, I had to run through the water to pass him round the outside of an island of stones. Disaster often seemed imminent, but never quite occurred; and he was gaffed at the tail of the island-a perfect fish of 11 lb.

The flies should be strongly tied with plenty of very stiff tackle and well oiled before use. They do not last long, and it is poor economy to fish with a draggled fly.

The hooks must be above reproach, and the gut strong, but it is advisable to have a somewhat finer point at the end of the cast during the day. It is most important to grease the casting-line thoroughly to ensure its floating. If a fly will not float properly and persists in lying on its back or side, it is best not to fish any more water until the fault is corrected. I have had a great deal of trouble with the annoying habit which some flies have of lighting on their backs; and, though many theories were started as to its cause, no satisfactory conclusion was reached. Sometimes it seemed to be a loop of gut, sometimes the balance of the fly itself, which caused the mischief. Personally, I suspect that it was more often the fisherman. Whatever the cause, it was no use moving on until the fly was made to float square on the water with both wings cheerfully cocked.

Owing to the pace of the stream, the slack line must be drawn in with the fingers after the cast has been made, in order to keep in touch with the fly and make the strike effective. It is no easy matter to regulate this shortening of the line in quick water; and I lost many fish through having too much slack when the fish rose. To strike hard, and yet not to be able (owing to the amount of slack) to produce any effect on the fly, is most annoying. Another frequent cause of annoyance is striking with a hook previously broken on the rocks behind one's back. To feel the hook go into a fish and yet not hold him, and then to find that you have been fishing with a thing no more dangerous to the trout than a buttonhook, is too much for one's equanimity. I used the same rod that I use in England for dry-fly fishing-a split cane, 9 feet 6 inches in length, and very light, without a steel centre. The reel took 35 yards of line and 100 yards of thin backing. This rod sounds an insignificant weapon for dealing with heavy fish, but it is surprising how quickly they can be killed, if the angler takes care to fight always with the stream on his side. At any rate, the smaller the rod the greater the satisfaction in killing a fish. It must be allowed, however, that some of my companions preferred a rod with a steel centre; but split cane is good enough, without the help of steel. Moreover, the extra weight is a bar to delicate casting, and delicate casting is essential.

It is a golden rule, not always easy to observe, never to let a fish get one foot below you. If this rule is kept, the very pace and violence of the water drowns the fish; and, generally speaking, it need not take more than about five minutes to kill even those of 10 lb. The fight, however, often takes one several hundred yards below where the fish has been hooked, and is as varied and quick as anyone could wish. The play of a sea trout is savage, violent and exhilarating, but there is not much method about his madness. He shakes the fly as a terrier shakes a rat. His amazing leaps and rushes, though exhausting to himself in heavy water, make his play more exciting to me than that of any other fish. The rod bends to an alarming curve and tingles with the strain. It has the effect of making me laugh, and I feel more confidence and less abject fear than when playing a brown trout. My confidence has no justification-I lose quite as many sea trout as brown trout, and far too many of both.

Wading is necessary in order that the angler may be as much as possible below the fish over which he is casting, and that his line may not cause the fly to drag. When the trout is hooked, the bank must be reached at once in preparation for a rush down-stream. Large sea trout leave the pool very readily; and a man must be ready to leave it, if possible, with some yards' start along the bank. 'Get away as near him as you can, and keep there,' is as good advice in catching sea trout as it is in catching foxes. For a rocky river the banks of this particular stream are fairly good going; but waders are not ideal things to run in, and fast running is necessary to keep ahead of the fish in the rapids. Once out of the pool, the fish is carried down by the stream tail first and near the surface. He obviously feels helpless, and is liable to get choked in the waves. When this happens he may often be gaffed very quickly, as he comes into the stiller water under the bank to try to get his wind. If the fisherman can manage to leave the pool ahead of the fish on a short line, he may feel pretty confident of killing him. If the fish gets out of the pool in front of him, he is likely to be lost, owing to the drowned line getting round a rock. Even if this does not happen, I always expect the hook to be torn out of