118 
BULLETIN OF THE UNITED STATES FISH COMMISSION. 
The lake is fed by a number of cascades, several of them very beautiful. There 
is absolutely uo stream flowing in over a bed in which fish cau spawn. The cascades 
either tumble directly into the lake or by a series of short waterfalls. At the head 
of the lake on the eastern side are the several dry, bowldery beds of a torrent, previ- 
ously alluded to, which form a junction about a quarter of a mile from the lake, and 
then continue as one bed, except where the stream in its flood has met some obstruc- 
tion and has torn its way througli by several channels, carrying everything before it. 
About half a mile from the lake a considerable flow of water falls over a log jam 
into a pool, where it is swallowed up. Advancing farther, the stream increases in 
volume and the mountains come close together, until the bed is hemmed in by a nar- 
row gorge, and one looks up vertically to the top of the 
mountain. In cue place there is a sheer fall of 800 to 1,000 
feet. About a mile from the lake the gorge ends, and the 
stream falls over the comparatively low gap in a beautiful 
cascade about 150 feet high. The rise from the lake to the 
cascade is about 1 to 10. The question is. What becomes 
of the water? On examination it is seen that huge slides 
from the surrounding mountains are not infrequent, and that 
the timber bordering the bed of the torrent grows uiiou the 
detritus; in fact, the whole shelf at the upper end of the 
lake is of the same material — great angular blocks, grading 
down to finer material. Great trees^ torn up by the roots 
and piled in with rocks carried down, are scattered every- 
where; bowlders piled up in fantastic shapes — all show the 
violence of tbe torrent. Trees that have been thrown across 
the stream and jammed have great hollows pounded in the 
upper surfaces by the rocks as they are swept onward by the 
current. This shelf, then, consists of these angular blocks; 
they are not waterworn, and there is no large amount of fine 
material. This formation probably extends to a considerable 
depth, and in the bed of the stream it iiermits the water to 
percolate through the interstices and find its way by seepage 
into the lake. When the stream is at its flood the volume is 
too great to sink away, and it flows into the lake as a torrent 
over what was a dry bed at the time of our visit. 
Under such conditions the redfish must spawn in the lake, and the cannery foreman, 
who winters here, states that he has frequently observed them spawning on the shelf off 
the dry bed of the torrent, making no nests. This shelf is poorly adapted for spawning- 
beds. It is only a few yards wide when it drops off into deep water, and is composed 
of rocky material that could not be used for nesting. The peculiar conditions that 
prevail here in reference to these spawning-grounds are exceedingly interesting, from 
the fact that redfish not only spawn in the lake, but under conditions entirely different 
from those usually believed to be necessary. Yet the stream can be relied upon for 
redfish, and a large number are taken very year in the vicinity of its mouth, or on the 
fishing-ground; the average catch during seven years was 34,903, the largest num- 
ber, 69,553, being taken in 1894. At the point selected for spawning-beds the salmon 
