We were returning from a long journey up the river, and had run out of provisions altogether. One of the men whom I had set to work to catch something, somehow, threw his bait into this hole casually, on his way down to the foot of the fall, and had a rise from a large fish. As anxious a day's fishing as I ever did succeeded this. A wary cast of a Jock Scott brought a fish to look at the fly, and turn back deliberately. After a half-hour interval he came again. Every fly in the book, and every dodge I knew, were pitted against the provoking indisposition of those Wananishe to be caught. At last it became a matter of personal pride as well as of hunger. Eventually artfulness and patience triumphed, and an interesting discovery, as well as a good supper, resulted; but it was hard to take measurements and notes of those fish before handing them to the cook.

The size of the Land-locked Salmon varies a good deal, in different waters, but is pretty uniform in each locality. According to Mr. Atkins, the Sebago and Union fish are larger than those of the Sebec and St. Croix. The Sebago fish average at spawning-time four or five pounds for the males, and a pound less for the females; but specimens running as high as twelve or fourteen pounds are not rare, and there is a record of one weighing seventeen and one-half pounds. The Union River fish are about the same size as those of Sebago.

The St. Croix fish vary in size at different parts of the water system in which they are found; those of the Schoodic River and Grand Lake Stream, where they are most numerous and where the hatchery is situated, average a little less than three pounds; specimens over six pounds are rare, and there is no record of anything over ten pounds. It appears that the average size in Grand Lake Stream increased to about four pounds between 1875 and 1884. These fish, however, were the ones specially taken for breeding purposes by Mr. Atkins, and therefore would probably be large; and owing to their protection for nine years, there would naturally be a greater number to select from.

The Lake Wenern Salmon of Sweden, like the larger of the Sebago examples, are equal to sea Salmon in size. Dr. Day gives the lengths of a couple examined by him as thirty-one and thirty-three inches, and other accounts show that they run from seven to twenty pounds.

The Nova Scotia and New Brunswick fish are small-the latter especially; a couple of pounds is a good weight for them, but the waters in which they occur are comparatively restricted in area, and they are much fished. Hallock states that he has seen specimens of the Salmon from the Stony Lake Chain, in Ontario, weighing twenty pounds. This-curiously enough, for it is the nearest to me-is the only one of the "Land-locked Salmon," besides the Swedish variety, that I have not personally examined; but I have never been able to visit the region during the fishing season, and cannot succeed in getting a specimen. I sometimes think Mr. Hallock's fish must have been the true Salmon of Lake Ontario, now all but extinct. In spite of efforts to preserve them, and to propagate them by artificial breeding, which promised for a time to be successful, the changed conditions of the streams, owing to the clearing and settlement of the country, have been fatal.

It is worth noting that, though game and game fishes can and do survive civilization in Europe, they soon disappear on this continent, under conditions that seem more favorable. Trout and Salmon manage to exist in British streams bristling with nets, weirs, dams, and all manner of destructive engines, and polluted by sewage and the refuse of manufactories, but in America, once the country is cultivated, they vanish. The preservation of their spawning-grounds in the natural condition, if they can get there at all, is probably the reason in the former case. In England the change in rain-fall, due to the disappearance of forests, can never have been so great as it is in America.

The Lake St. John and Saguenay fish average a little over two and one-half pounds. Four-pound fish were numerous enough a few years ago, but anything over that size is large, and only occasionally will a six-pounder be found. Out of many thousands I have seen but one seven-pound fish; it was twenty-seven inches in length, and a very lank specimen. If properly filled out, it would have weighed nine or ten pounds. This solitary instance gives one some faith in the stories of the large size of the Wananishe when the region was first settled, forty years ago. Occasionally very large ones are seen feeding by themselves, but they are extremely wary, and there is no authentic record of one above seven pounds, though the late Senator David Price, of Chicoutimi, is said to have caught one of eleven pounds in weight.

I did not get enough of the Labrador fish to establish an average, but I imagine them to be large, because of the abundance of food, great area of the lakes, and freedom from disturbance. My specimens varied from a quarter of a pound to six and one-half pounds. The Indians said much larger fish were plentiful far up the rivers, but we all know how that is ourselves.

As my own observations have been chiefly of the Wananishe, I will confine myself to the appearance and dimensions of this variety, which agree very closely with those of the Schoodic Salmon as described by Mr. Atkins. The Wananishe is a much longer fish, and altogether larger, in proportion to his weight, than is the Brook Trout, as the following figures show:

Specimens under a pound are rare, and are found in shallow water and small streams. It was a long time before I succeeded in getting a Wananishe parr, even in water which I knew to be just below favorite spawning-grounds. They are almost indistinguishable from Salmon parr, and are probably taken for small Trout, if ever observed at all. The four-ounce fish have already put on the silvery livery of the smolt, through which the transverse bands of the parr-marking show distinctly. In specimens of from a pound to a pound and one-quarter the silver scales rub off easily, and the parr-bands are to be seen even on fish up to two pounds in weight. Mr. Atkins states that the marks are distinct on the under side of the skin of adult fish of the Schoodic variety. This persistency of the parr-marking is considered by him to be evidence of arrested development, and perhaps rightly so, when we connect with it the fact, established by the Howietoun experiments, that the parr and smolt of the sea Salmon, both male and female, when bred and raised entirely in fresh water, can reproduce their species, and that their progeny again are fertile.