Orvis rod half an hour's good work. It measured twenty-one inches, but was very light for the length, weighing but two and three-quarter pounds. At the Redoubt river, much larger ones were taken; and two which I shot in Beardslee river were over two feet in length; how much they weighed I never found out, for their surroundings of sick and dying salmon, upon whose eggs they were feeding, prejudiced me against them and I left them.

In shape and color the spectabilis vary greatly, both factors depending upon the length of time they have been in fresh water. When fresh run, they are long and lean, shaped somewhat like the lake trout of Adirondack lakes. The colors are dark lustrous olive-green back, growing lighter as the median line is approached, and blending into a silvery gray tint, which pales to a pure white on the belly; the green portion is sprinkled with golden specks; the flesh is hard, and very good for the table. After a yery short sojourn in the creek, bright crimson specks appear among the golden, which, however, fade to a pale yellow; the lustre of the green disappears, they become heavier, but the flesh becomes soft and uneatable, and the skin is covered with slime. Salmon trout taken late in August and early in September, were full of ripe ova.

Professor Bean placed some fish, that had been taken in salt water, into a bucket of fresh, and the crimson spots made their appearance in less than a day.

When fully decked with these, and fattened, they resembled our fontanalis greatly—the head, however, being somewhat larger, and the tail less square.

Salmo Gardneri. My acquaintance with this species is yery limited. The first one that I saw I took in Sawmill Creek, well up to the head, in September, 1879. Seeing that it differed greatly from the spectabilis, I preserved it in alcohol, and it was subsequently identified by Professor Bean. It measured a trifle over ten inches, and was very plump, weighing seven and a quarter ounces. In my notes, I describe it thus: "Body, dark green on back, but in general colors very much like a steel head or quinnat salmon; covered with round, black spots, from one-sixteenth to one-eighth inch in diameter; these extend considerably below the median line, and the tail and dorsal fins are covered with them; the second dorsal adipose, but less so than that of the fontanalis, having a slight show of membrane, on which there are four spots; ventral and anal fins, yellowish in centre, bordered with red; belly, dull white; tail, nearly square; scales, quite large, about the size of those of a fingerling chub; flesh, firm; and skin, not slimy. No signs of ova or milt."

On the 28th of April, 1880, I made note: "The first salmon of the season made their début to-day—that is, if they are salmon, which I doubt.

"Five beauties, from thirty to forty inches long, were brought alongside, in a canoe paddled by a wild-looking and awe-struck Siwash, who, with his crouching Klootchman and papoose, gazed upon our ship, guns, and us with an expression that showed them to be unfamiliar sights. He was evidently a stranger, and was taken in, for he took willingly two bits (25 cents) each for the fish, and no Sitka Siwasli but would have charged treble the price. Through an interpreter, I learned that he had spent the last seven months in a shanty on the western side of Kruzoff Island, and that well up, among the foot-hills of Mount Edgecomb, there was a little lake, from which there flowed a small stream into the Pacific, and that in the headwaters of this stream he had speared these fish, which run up the stream in the fall, remain all winter in the lake, and in early spring spawn in the head of the outlet."

All of this militated strongly against the theory that they were salmon, and when, on being dressed, the females were found to be full of ripe ova, said theory was upset completely. My ten-inch specimen of last September supplied us with a clue, and it was soon decided that these magnificent fish were indeed trout; for in every respect except size, and size of spots, some of which were a quarter of an inch in diameter, the fish were identical. Whitford, the oldest inhabitant, confirmed the Indian's story, and gave me in addition the Indian name for the fish—Quot and that of the Russians, which I forget, but it meant "Mountain Trout," and said that they are found only in the lakes, high up in the mountains, and that in winter the Indians spear and catch them through holes in the ice.

We found the flesh to be very delicious—far more so than the best of the salmon. The processes of cooking, both by broiling and boiling, had a curious effect, for the flesh, which, when uncooked, was of a very bright red, blanched to pure white.

The trip to Mount Edgecomb, in the early spring, involved hardship and danger; and although several of us resolved that we would undertake it, for the sake of such fish, somehow we never did, and I have thus described all of the gardneri that I ever saw.

Salmo purpuratus (Clarkii). The most beautiful of the trout family, although in no way equal to our Eastern trout in any other respect.

The purpuratus is a lake trout, and found only in low-lying lakes. Just back of Sitka, at the foot of the mountains, and elevated perhaps twenty feet above the sea, is a little lake dubbed by me "Piseco".

Handy to get at, and its outlet running through the centre of the town, it became, in early spring, our first resort for fishing. Arriving in June, 1879, many of us had, through days of fruitless endeavor, during the summer and autumn, grown to disbelieve the tales of the inhabitants, that this lake abounded in trout; but on the 20th of May, 1880, from somewhere, there thronged the shallow edges, among the lily pads, great schools of these trout, and for about two weeks there was no limit to the number one could take of them. Salmon spawn was the best bait, but a bit of venison would answer. A fly they would not rise to. In size, they ranged from six to twelve inches—the latter size being, however, very exceptional; their average was about eight inches. The description in my notes is: "Specimen, May 27th. Length, nine and one-half inches; depth, two and three-eighth inches; weight, five ounces; colors—back, rich, dark brown, growing lighter toward medial line; at which, covering it for a space of half an inch, there is a longitudinal stripe of rich purple, extending from opercle nearly to tail; below the median line, bright olive-green, lightening to silvery white on belly. All of the tinted portion is profusely sprinkled with oval black spots, which mark also the dorsal, caudal, and adipose fins; the ventral and anal fins are yellowish bordered with crimson; tail, nearly square.

"The entire tinted portion has a beautiful golden iridescence, so that when held in the sunlight, and looked at from the rear, it seems to be gilded." It may be noticed that, with the exception of the purple stripe and the golden iridescence, the description of this fish is almost identical with that of the gardneri. I think it quite possible that they are the same at different ages, and that later in life these Clarkii may become ambitious and seek more lofty lakes. None that were taken contained ova.

Where they came from, unless they run up the inlet at night, no one found out, for although closely watched in the daytime, none were ever seen in it.

After about two weeks the greater portion disappeared, and although sought in the deep waters of the lake, could not be found. Major William Governeur Morris, the Collector of Customs of Alaska, assures me, however, that during the summer of 1882, he found certain places in the lake where he caught them until August. On July 4th he with a friend catching four hundred and three in three hours, baiting with a single salmon egg.

I am not sure that we could not have again found them, but the fishing having grown slack in the lake, and growing daily better in the creeks, we spent most of our time