CXII REPORT OF COMMISSIONER OF FISH AND FISHERIES. 



a northeast or northerly direction. Those tributary to the Clarke Fork of the 

 Columbia flow to the northwest, Avhilo the drainage into the Snake River or Lewis 

 Fork of the Columbia is to the southwest. 



Nearly all of these rivers and creeks are, of course, swift mountain streams; most 

 of them have their rise in small lakes of clear, cold water, high up in the mountains — 

 lakes which as yet are ditlicnlt of access and but little known. Many of these lakes 

 are known, however, to be well supplied with trout, while others are wholly with- 

 out any lish life whatever. From these mountain lakes the swiftly-tiowing, turbu- 

 lent streams make their descent through rocky gorges and canyons to the valleys 

 below. Ordinarily the beds of the streams are very rocky, but now and then are 

 found more (juiet reaches where the streams have sand or gravel beds as they flow 

 through small mountain meadows. Then, at other places, there are rajjids and cas- 

 cades, and in many of the streams are found considerable waterJ'alls. The best 

 illustrations of this are in the numerous magnificent falls found in the streams 

 flowing from the great volcanic plateau constituting the larger part of the Yellow- 

 stone National Park. As the streams leave this immense sheet of rhyolite they do so 

 in great falls, such as those of the Yellowstone, Gibbon, and Lewis rivers. Others 

 of the same nature are to be found in the country lying to the east of the National 

 Park, in the Clarke Fork of the Yellowstone, and other streams of that region. These 

 falls, of course, serve as more or less efiective barriers to the distribution of fish, 

 and as a result many of the mountain lakes, though of the most suitable character, 

 so far as temperatTire, purity, and abundance of food supply are concerned, are 

 wholly barren of lish life of any kind. The larger streams are, as a rule, less swift, 

 and have more sandy and gravelly beds. There are few, if any, sluggish streams in 

 this region, and all are clear, unless contaminated by mining operations. 



In the Columbia Kiver basin twenty-four lakes and streams tributary 

 to tlie Clarke Fork and nine tributary to the Lewis Fork were examined 

 by Prof. Evermanu; in the Missouri River basin the examinations 

 extended to twenty-six tributaries of the Yellowstone River and Lake, 

 the Madison, Gallatin, and Jefferson rivers, and Prickly Pear Creek. 

 Descriptions are given of the jirincipal features characteristic of each 

 of these waters, and their fishes are discussed. 



One of the most interesting incidents of the expedition was a trip to 

 Two Ocean Pass,where the waters of the Columbia and Missouri rivers 

 virtually meet and provide a limited passageway for fishes from one to 

 the other. Visits had previously been paid to this locality, which is 

 situated just south of the southern border of the Yellowstone National 

 Park, by a few travelers who have given accounts of its peculiarities, 

 but although Prof. Evermann was there only a short tin)e his observa- 

 tions have enabled him to explain its features more completely and 

 accurately than any of his predecessors, whose descriptions of the place 

 are greatly at variance with one another. His report contains a plan 

 and sketch illustrating the true relations existing between the several 

 sti'eams. 



According to Prof. Evermann — 



Two-Ocean Pass is a nearly level piece of meadow land, surrounded by rather high 

 hills except where the narrow valleys of Atlantic and Pacific creeks open out from it. 

 Ruuning l)ack from the hills to the northward are two small canyons ; on the opposite 

 side is another canyon of the same character. Down these canyons come the three 

 main streams which flow through the pass. The extreme length of the pass from 

 east to west can not be mu<'h less than a mile, while the width from north to south 

 is perhaps three-fourths of a mile. 



