When one sees a number of salmon side by side, as they often lie in the tail of a pool, and watches a regular salmon fly pass over them or past their very noses without any attention or motion on their part, except to move away if the fly or leader comes too close, one is tempted to wonder if these fish will really take a fly at all under these conditions. And yet it is these very fish at this time which will furnish the best of sport. For some reason they are in a state of mind where the wet fly does not attract them at all. Perhaps they have reverted to the state when they were parr, taking insects off the surface. Let a real fly or a small butterfly float over them and see how often one will rise and suck it in. It was observing this which made me try a dry fly, with not much success at first, because I did not know how to use it, but soon I made a proper cast, quite by accident, and raised a fish. The fly was a Greenwell's Glory, No. 14 hook. I soon observed that the fish rose on some kinds of casts, but never on others, with considerable regularity. I was interested to discover which casts were wrong. The place was on the Indian River in Newfoundland, at the outlet of a lake where there were a large number of salmon. For a week previously they had bitten well and we had had good sport, but suddenly overnight the fishing had stopped. Conditions had changed. The water temperature had risen to above 60 deg. F. The fish were there, I could see them, so there was no use going anywhere else. If I could not get them there I had little chance elsewhere, so I settled down to find out how to catch them.

The first thing to discover was what effect the leader and its shadow had on the fish. To discern this I tied leaders of various sizes, going down to the finest gut made (.004-inch diameter) and up to the regular salmon gut (.020-inch diameter). I always carry a small gauge for measuring the size in order to make the leader of the proper taper so it will cast well. I soon found when the sun was out I raised many more fish with fine gut, and those which did come up almost invariably got the fly, while those which raised on the coarse gut very often turned away just before taking the fly in their mouth; they seemed to see the leader and be frightened.

The problem became largely one of fishing so that they would not see the leader. This can be done in two ways, first by having it as thin as possible. This is not very practical because it is difficult to hook and land a fish of considerable size on a gut which will break under a strain of a pound and a half or less. The second way is to see if the leader could not be cast so that they would not see it is readily. I soon found this to be possible to a certain extent. If the leader extends up stream from the salmon for a couple of feet or more it is in a straight line away from him when he comes to take the fly, and he is not so likely to see it in time to avoid the fly. This is the key to dry fly fishing for salmon. Have the fly float directly over the fish so that he will see it with both eyes and have the leader lead directly away from him. If the fly was pulled on the surface, I found that he was far less likely to rise and a fly partly submerged almost never takes a fish. It must float on the surface of the water well up on the top of its hackles.

The light effects are what attract the fish, because this is what occurs when natural insects float on the surface. Be this as it may, the fact remains that to be successful you must either use the fly on top or submerged. A dry fly pulled below the surface often works well. It seems as if the fish saw the fly on the surface and did not take it but decided to do so when it was being pulled away below the surface. Under some conditions this method takes fish better than any other. One afternoon recently I hooked fifteen fish nearly all in this way and found it better than a dry or wet fly used alone.

In regular fishing I find that too many fish are lost by breaking the leader in hooking them if the size of the leader near the fly is less than .010. I have gut of this size which pulls four pounds, but this is very exceptional. The usual gut pulls about two to two and one-half pounds in this size. When I find that fish are not being hooked and are missing the fly I put on smaller gut. In order not to break this fine tackle, I have had made a special form of rod which is adapted to this purpose. It is ten feet six inches long, the greatest length which can be continuously used without undue fatigue in one hand. The weight is 7 oz. Without the extra hand-piece below the reel which I use in order not to become too tired in my left wrist if I hook a large fish and have to play him a long time. I can rest the butt of the rod against my body, and get at the reel with my right hand without danger of getting the handle tangled in my clothes. Such a rod will not pull over one pound when the tip is up and the greatest strain possible put on the butt. This will not stop or turn a good sized salmon but will tire him out quickly. Of course, the pleasure of playing a salmon on good heavy tackle and giving him the butt and making him jump is sacrificed to the pleasure of seeing him rise to the surface and hooking him, often at close range. But with this tackle I raise and hook many more fish. Such a rod works best with a tapered E line about .035-inch diameter in the heavy part. If it is well handled it will put a fly ninety feet and lay it down lightly. I have hooked fish at this distance quite often and regularly fish at seventy feet with ease. The leader must be very long for dry fly work, as I find the line floating over the fish or striking the water scares them badly. I use regularly a fourteen-foot leader with the large end about .020-inch diameter. Often I have found that this is not long enough and added three to six feet more finer gut.

It is very hard for the novice to tell when the salmon is alarmed. He does not usually run away, as does the trout, unless he is very frightened indeed. If you watch carefully the only thing you can observe is that he settles a little toward the bottom and often changes the motion of his fins and tail. When this happens a salmon will not rise and there is no use casting over him just then. On the other hand if he begins to work his front fins faster and raises his head or whole body in the water as the fly passes over him, he is taking notice and will most surely rise to the fly if it is put over him enough times in the proper way. When I have seen him take notice I have sometimes kept at him an hour or so, and almost invariably raised him to the fly. This is quite the opposite from the effect of a wet fly, which offers its greatest attraction when first seen, and the probability that he will take it decreases with the number of times the salmon sees it. It is for this reason that the old fishermen rest a fish which has risen short and are certain that they have more chance of taking him if he gets a fresh view of the fly after an interval. I feel that the warm water has in some mysterious way brought to the surface of consciousness the habits formed during the parr stage of growth of taking insects on the surface, and that the adult salmon behaves in a similar way, driven on by these old habits which became his dominant impulse because he is cut off from his regular feeding on fish, squid, etc.,

It is on this theory that I have worked in studying the dry fly and the remarkable results which I will describe later seem to bear out my conception. Dry fly fishing seems to be regulated by the water temperature and the lateness of the season. Until this year I had always supposed that dry fly fishing was better the later it was practiced; but this year about August 2 I found that in some places salmon would not take a dry fly when they had taken it readily July 12-15. This may be due to the fact that this season was very exceptional, and that the fish were much further advanced toward spawning than usual. In other years I have always had the best of dry fly fishing on August 15, and have had excellent sport in Newfoundland in September.

It has been my experience that salmon do not take a dry fly well when the water is below 58 deg. F. and begin to take it well when the water is 60 deg. F. The best dry fly condition is with the water 60 deg.-66 deg. F. It may not be generally known, but the long Northern days warm up salmon rivers greatly before night. On this trip I frequently noticed a rise of 10 degrees from 8 a.m. to 5 p.m. The size of the flies seem to be very important; at times the small flies, even as small as No. 16, seem to be better, while at other times very large flies as big as two-inch in diameter seem to raise more fish. It perhaps somewhat depends on the depth of the fish in the water but more probably on the clearness and light at the time or the background in view from the window of the fish.

I remember one day fishing a run on the Upsalquitch with Mr. Monell. The fish were in about two or three feet of water and under a moderate current running rather smooth. The bottom was covered with stones of about a foot to two feet in diameter. I was using a small gray hackle of about three-quarters of an inch in diameter. He was using a larger gray hackle of the same pattern, an inch and a quarter to an inch and a half in diameter. We took in all about forty fish in the afternoon. Those I got ran from six to eight pounds, while those he hooked ran from ten to twelve pounds. We did this in order to see whether the larger fly took the larger fish, and it certainly proved so in this case. Recently in the Restigouche I could only raise one fish on the large inch and a half hackle, and got as many as I wanted on the smaller three-quarter inch hackle. These fish ran from about twelve to eighteen pounds. Another time at The Forks on the Upsalquitch on a clear, hot afternoon I could not raise any from a bunch of fish along a ledge with the three-quarter inch or the one and a half inch diameter flies. I went into camp and tied the largest hackle I could, over to and a quarter inches in diameter, and immediately hooked three fish in succession. It will be seen from this that no general rule can be given for the fly to use at all times; a man must be a fisherman to know. He should experiment and find out the best size for each day.

Recently I had a most interesting example of how the diameter of a leader affects the number of fish hooked. I was fishing the run above Jimmy's Hole on the Restigouche and took one salmon from the fifteen or twenty spread across the bar at the top, in about two feet of water. The leader was twelve feet long and .014-inch diameter at the small end. I raised about twenty fish in succession, but failed to hook a single one. They all seemed to miss the fly. After spending about three-quarters of an hour in this way, to make sure of the matter, I put on a thin end about .010-inch diameter and four feet long. This did two things: it made the leader less visible near the fly and it enabled me to cast the fly more gently on the water so that it floated up on the hackles. The result was remarkable; I hooked seven fish in succession, missing only a few of those which rose. This experiment satisfied me that my tackle had not been right in the first place. Another time at the Nine Mile Pool on the N.W. branch of the Upsalquitch on a very bright, clear, hot day, we failed to get any fish in the run at the top but could see a large number about three feet below the surface in the center of the pool in deep water. I put on a dry fly with a .010-inch leader and failed to raise any fish at all; I then got out a very fine drawn gut leader about nine feet long and .005-inch diameter, and added it to the other leader, making it all about twenty feet, and used the same fly, but took a very light trout rod so as not to break the fine gut. On the second cast I hooked a fish, and by holding very gently finally made him come down to the gaff in the shallow water at the tail of the pool. In this way I hooked and landed five, none over ten pounds or under eight pounds. A scow with horses then came through the pool, and scared the fish so that I could not get any more before we were obliged to leave. I was convinced that I could not have raised any at all in this pool with any other ordinary leader. I never use a leader any finer than necessary, but always try very fine tackle if fish will not raise with my regular leaders. Of course I lose a lot of fish, but don't I have a good time doing it! Recently on the Spring Pool of the upper Restigouche I lost five fish in succession and then two more on the next pool, yet with all that I brought home two fish, which is two more than I should have had if I had not used fine tackle; besides I had a splendid day in place of a complete blank. The important thing to remember in dry fly fishing is that the fly must pass directly over the fish. Near him will not do any good at all. For this reason the angler must know just where the fish is. This is generally accomplished by looking the pool over and seeing just where the salmon are located. It is an easy matter to do this with a canoe without unduly scaring the fish, and it often saves a lot of useless casting. Recently I carefully fished a ledgy pool where I had seen twenty-five fish the evening before, getting only one rise at the tail; the rest had moved on in the night. Sometimes the current makes past a jutting point of rock behind the corner of which is a sure place for salmon, if there is one in the pool. Such a place is always worth trying without looking, because if scared from there the fish may not return for some time, and if he is there, he is almost sure to rise.

Along ledges is a favorite place for dry fly work because the fish are sure to lie in certain positions if there and the fly can be easily made to pass right over them. But it is at the trail of the pool that I love most to see them or in places where spring water comes in, because they are in plain sight and I can generally place myself in such a position that I can see them without alarming them at all. Very often I can get within thirty or forty feet of them and witness the whole performance. It happened often that Mr. La Branche on one occasion found seven fish together and got six of them before dark. On another occasion I found three fish in about two feet of water and finally got all three, but as a general rule I expect to hook from ten to twenty per cent of the fish I cast over in this way. It is very rare that I cannot raise any of them, and then only when it is very hot and clear in the middle of the day.

We will now transport ourselves to a salmon river in late July or early August and see just what will happen. The time of day will be 11 a.m.; the pool, a long one with a ripply run in it at the top and a deep water center and a tail end with the bottom of stones of the size of your hand to two feet in diameter.

We come to the pool at the top and I stand up in the canoe and let the guide paddle carefully, not using his steel-shod pole, which scares the fish. We drift down the side of the current at the top and see no sign of fish. In the center of the pool are several salmon in deep water along ledges; as the canoe nears the tail we draw to one side and look carefully in a small depression caused by an ice or log jam. Here are six salmon with their noses pointed up stream. The water is not over three feet deep at most and getting gradually shallower toward the land. We carefully back the canoe and pull it up out of the way and the guide climbs up on a projecting log where he can see the fish, and tell me if the fly passes over them and what they do. I get out my 10 foot 6 inch rod with a fourteen-foot leader and a gray hackle fly and wade into position. I always fish on my feet and wade if possible. A canoe is very hard to fish from with a dry fly because it scares the fish and because it is necessary to make a disturbance when you move. In dry fly fishing the position of the fly, the leader, and the line are all of them important, and it is hard to judge where to place yourself to get a perfect cast--taking into consideration the wind and the current, which make a difference. I always start in well below the fish and to one side, so that they do not see me at all.

My position will be about forty feet to the right of the fish and perhaps fifteen feet below them. I get out my line, casting in the air and up stream, along the bank, until I judge that I have the right amount of line out, so that the fly will light three or four feet upstream from the nearest fish and directly in line with him. The fly is well oiled with a mixture of albolene and kerosene in equal parts and floats well; the leader and line are carefully greased with deer's fat and float on the surface. The fly is cast with a curl in the leader so that it floats right over the fish; this is best done by shooting the line through guides and checking it with the left hand before the fly lights. This jerks the fly back and causes a curl. A little practice will soon teach the trick. The salmon seems to pay no attention to the fly, but his head rises visibly from the bottom. A second cast does not come so close over him and the guide says the fly is too near to me. I lengthen out the casts a little and place the fly just right, about two feet in front of him; as it floats down over him I see him rise and come rather slowly to the surface. As his head comes up I hear the sucking noise which is made by closing the gill plates and suddenly opening the mouth, causing the fly to enter. I have been fishing trout too much this year and strike quickly and pull the fly away before the fish gets it.

It is bad work, and I have to take my punishment by waiting until the salmon resumes his position in the group. They generally take up almost the same position as before. I begin my casting again, and in a few casts the guide says, "the fly is passing just over him." This time he does not rise directly up as before but turns after the fly has passed and gets below it, rises and takes it with a great rush. There was no missing this rise, and a lifting of the line sets the fly and the fight is on. He runs a hundred feet or so and jumps into the air about six feet clear of the water, tumbling over directly away from me. This kind of leap is very likely to lose the salmon, as he almost invariably hits the leader with the tail when he jumps. There is a conviction among guides that the salmon always hits the leader and that the salmon causes splits in his tail by so doing. The number of splits in his tail corresponds to the number of jumps. I have generally found this to be the case but it may be only a coincidence. The salmon runs up to the head of the pool and then down to the deep water where he begins to "chug," as it is called; this is jerking the head against the pulls of the line to loosen the hook. Salmon generally do this when they are lightly hooked and generally get off. I always dislike to feel it. The only thing to do is to hold them very lightly so they will not have much strain to work against. They generally soon stop and begin to run again if they don't tear out. The fish jumps a second time and makes for the end of the pool with the evident intention of going down stream. I ease the strain on him as much as possible and run along the bank and get below him; this turns him upstream, as they generally fight away from the strain on the line. Bearing right to the edge of the swift water he turns and goes up into the pool and makes runs, getting shorter and shorter as the strain of the line and fighting the current gradually tire him out.

With the light rod it often takes quite a time to get a fish close enough to gaff; he makes many short runs as he sees the guide, but the pull of the line gradually brings him in and at last he is landed on the beach, a fine fifteen-pound fish. This is all right for a starter, but there are five more fish there. We look and see that the disturbance of the pool has not caused them to move. Again I get into position for another cast and put the fly over the next nearest fish. Twenty casts or so fail to make him move in the least, so I cast a little further over to the next two fish, which are almost in line with each other; the second one moves upward; here is another chance for a rise. The fly lights only a few inches in front of his nose. He turns his head upward and instead of making a turn to take the fly he raises his head vertically upward and pushes his whole body out of the water as far as the back fin with the fly in his mouth. The strike pulls him over and he seems astonished as he jumps at once four or five feet clear of the water, a fine fourteen pound fish, and off he goes again across the pool directly over the bunch of fish I had been fishing for. This makes them restless and they take up entirely new positions. While playing my fish on the line I watched them and saw one of the fish I had noticed in the deep water swim slowly down out of the pool and swing into line like a cavalryman taking his position in the ranks.

It often happens while fishing the pool that more fish come into view from the deep water. One never knows what a salmon will do next. This one just ran out of the pool and down the rapids so that I had to call for the canoe, as he already had out over 400 feet of my 600 feet of line. I got in and reeled up as the canoe rapidly overtook him. We soon passed him and downstream caused him to stop and turn up. The current soon tired him out and we brought him to gaff in a little eddy at the side.

The fishing of this pool is characteristic of dry fly fishing in July or August in low water.