The second form in which I mistook it for a distinct species is that of the Humpbacked Salmon (_Salmo proteus_, Pallas; _Salmo gibber_, Suckley; ‘_gerbuscha_,’ Kamtschatka; ‘_hud-do_’ of the Nesqually Indians; ‘_hun-num_’ of the Fraser river Indians). In its general outline it differs altogether from the Hook-nosed Salmon. The back is much more arched; nose curved, but not nearly as much as in the mature _Salmo lycaodon_, and the under-jaw turns up and terminates in a protuberance or knob; teeth much more numerous, sharper, and smaller; tail deeply notched, and thickly spotted with dark oval-shaped marks. The most conspicuous feature is a large hump of adipose material situated on the shoulders, a little anterior to the dorsal fin, and only found in the male fish. It has generally been stated that this hump grows upon the male fish after entering the fresh-water: this is a mistake, for I have seen them again and again taken in the sea, before going up into the rivers, with this hump well developed. On cutting it open, it appears to be a sort of cellular membrane, filled with an oily, semifluid kind of material. The use of this deposit, there can be no doubt, is to supply the male with this material in some mysterious way during the spawning-time, for, after that period has passed, the hump entirely disappears. They arrive about the same time as the older fish, but only in very large runs every second year—have the same range, and die in thousands. At Fort Hope, on the Fraser river, in the month of September, I was going trout-fishing in a beautiful stream, the Qua-que-alla, that comes thundering and dancing down the Cascade Mountains, cold and clear as crystal; these salmon were then toiling up in thousands, and were so thick in the ford that I had great trouble to ride my horse through; the salmon were in such numbers about his legs as to impede his progress, and frightened him so, that he plunged viciously and very nearly had me off. They are never at any time good eating; the flesh, in fresh-run fish, is white, soft, and tasteless. The Indians only eat them when they are unable to obtain anything else. These salmon work up to the very heads of the tributaries, and I have often seen them where the water was so shallow as to leave their backs uncovered. The _Salmo canis_ of Suckley (Dog-Salmon, Spotted Salmon, ‘Natural History of Washington Territory,’ p. 341), which he says arrives at Puget’s Sound in September and October, I believe to be only the old males of the _Salmo lycaodon_ (Hook-nosed Salmon), that have had a turn in the rivers perhaps a year or two before, and have got safely back again to the sea, recruited their wasted energies, and returned again for another perilous cruise up the streams. The large fanglike teeth, from which they derive the name of dog-salmon, are the large teeth grown and developed, as I have previously described them, whilst spawning in the fresh water.