RECREA TION. 



tactor in the permanence of new forms. 

 In Wab a lake* in Washington, a glacial 

 lake which has now no outlet, the trout 

 became entirely cut off from the parent 

 stock, and a local race, with shorter head 

 and the black spots gathered on the 

 tail, was formed by the separation. 



The trout thus came to the fountain 

 head of the Columbia, and its great 

 tributaries, the Snake, the Salmon and 

 Clark's Fork. How it crossed the Great 

 Divide over to the headwaters of the 

 Missouri and spread itself, where it 

 could, in the Yellowstone Park, I have 

 already twice told in my way. Dr. 

 Evermann, of the U. S. Fish Com- 

 mission, has told it in a still better way, 

 for he has himself visited the Two 

 Ocean pass and caught it in the act of 

 crossing the Divide. Just south of the 

 Yellowstone Park is a great depression 

 in the main divide of the Rocky Moun- 

 tain chain, which is reduced to a 

 quarter of a mile of low marshy ground. 

 East of this marsh the Atlantic creek 

 flows eastward into the Yellowstone. 

 West of it, Pacific creek finds its way 

 into Snake river. Across the marsh 

 the streams become entangled and each 

 one sends a part of its water across 

 into the other. In the spring the marsh 

 is largely under water and there is no 

 obstacle to the passage of the trout. 

 For the greater part of the year, one 

 stream, at least, is open, and the trout 

 can pass without hindrance from the 

 Snake river to the Yellowstone, from 

 the basin of the Columbia to that of 

 the Missouri.f 



Thus the trout came over into Yellow- 

 stone lake and into the Yellowstone 

 river, thence into the Missouri and its 

 great, clear affluents, the Jefferson, 

 Madison and Gallatin, and throughout 

 the Missouri basin as far to the east as 

 a decent fish can live. 



The wash of the Bad Lands in Da- 

 kota fills the river with fine clay 

 and quicksands ; and in yellow water 

 over quicksand bottom one does not 

 look for trout. The Black Hills of 

 South Dakota are full of clear streams, 



* The Waha Lake trout has received the name of 

 Sal-mo mykiss bouvieri, from Captain Bendire, its dis- 

 coverer; one soldier naming it for another. 



t The trout of the upper Missouri has been called 

 Sal-mo lewisi, by Girard, and Salmo carinatushy Cope. 

 It does not differ in any visible way from Salmo mykiss. 

 Trout confined to rivers are always smaller than those 

 of the same kind resident in lakes. Those which 

 enter the sea grow to a still larger size. 



but there are no trout in them. The 

 bad water of the main river into which 

 these streams flow shuts off the trout 

 from them. The fact that the trout 

 are shut out shows that conditions have 

 not materially changed since the trout 

 came into the Missouri. The cataracts 

 which fall from the lava beds in the 

 Yellowstone Park have also excluded 

 trout from a great number of beautiful 

 streams as the Gardiner, Gibbon and 

 Firehole rivers, and the charming ex- 

 panse of Shoshone* and Lewis lakes. 

 This shows that these waterfalls were 

 formed before the trout crossed the 

 Divide. 



From the tributaries of the Missouri 

 or the Snake, the trout crossed, in some 

 way as yet unknown, to the headwaters 

 of the Platte, and filled all the brook- 

 lets of the Colorado Parks. From these 

 it again overflowed into the neighbor- 

 ing waters of the upper Arkansas. The 

 fact that through ail these streams of 

 Colorado and Wyoming the trout are 

 substantially alike, indicates that the 

 date of crossing from one to the other, 

 say from Denver to Pueblo, is com- 

 paratively recent. The runway is not, 

 however, yet made out, but it probably 

 lies between Pikes Peak and Denver 

 and may have been due to some glacial 

 overflow, from the South Platte into the 

 creek called Font-qui-Bouille. The pass- 

 age from the Missouri to the Platte is 

 older, for here the trout have became 

 perceptibly changed. The trout of the 

 Platte f and Arkansas is small, very 

 green in color, with very red flesh ; the 

 spots are gathered chiefly on the tail and 

 the red cut throat mark is bright. 



From the Arkansas river to the Rio 

 Grande, over the Sangre de Cristo moun- 

 tains, is again but a step — a short step, 

 but a very high one. This again the trout 

 has, in some way, crossed. Here we 

 may imagine glacial lakes, now drained, 

 as the way of passage ; or still better 

 we may say we do not know. The 

 transfer must have been an old one, 

 for the trout in the Rio Grande \ is 

 visibly different, the variation lying in 



* In these streams are now trout in abundance, 

 various species having been introduced by the U. S. 

 Fish Commission, in 1888. 



tThe "Greenback trout" of the Arkansas and 

 Platte is Salmo mykiss stomias, Cope. 



X The trout of the Rio Grande is Salmo mykiss 

 spihirus, Cope. Its range extends farther southward 

 than that of any other known form ; as far as the 

 mountains of Chihuahua. 



