COPPER RIVER INVESTIGATION. 127 



After returning to the railrojid bridge a trip was niade upstream far 

 en'High to reach the point at which it emerged from the nari'ow 

 valley in the mountains where it originated. The stream was 

 admirably adapted to be the spawning ground for the salmon, and 

 its waters were already well occupied by fish approaching maturity. 

 Two hundred yards above the bridge was a group of about 100 fish 

 of all colors, a majority of which were well marked with red. Every 

 few yards farther another school was brought to light, varying in 

 size in a manner apparently well related to the area available in 

 that particular pool or curve of the stream. About a mile or less 

 above the bridge the stream flowed through a low flat area where its 

 gentle curn nt suggested lake conditions. It was slow and quiet as 

 well as warmer than the water in the stream below. This stretch 

 was shallow and bordered by rushes rnd sedges, and near it were 

 some little slougks communicating with the main creek where there 

 was an abimdance of green alga^ and some clumps of water grasses. 

 In these places large numbers of salmon fry were found. They were 

 chiefly sockeyes, though some young silver salmon were captured 

 also. Near the mouth of the canyon the valley was about one-half 

 mile wide, the stream winding with a more rapid flow as it descended 

 from the mountains. The topography of the countr}^ clearly showed 

 that there was no lake on the stream above that point, and there 

 the creek showed no evidence of the presence of salmon. In fact, 

 none were seen above the slow quiet stretch passing through the 

 boggy area. The temperature in the quiet stretch was from 1° to 2° 

 higher than in the water below, and the reason was soon found in 

 the presence of seepage areas and springs in the bottom of the stream 

 where water entered in sufficient quantities to make the volume of 

 flow near the railroad bridge two or three times that which was 

 present at the place where the stream passed through the swamp. 

 Clear Creek was visited again on September 6 wlien on the way back 

 to Cordova. Spawning salmon were still pres( nt in the water near 

 the railroad bridge. There was no opportunity at that time to 

 investigate them or the stream further. 



ABERCROMBIE CANYON AND RAPIDS. 



In view of the conflicting statements concerning the actions of the 

 salmon ascending this part of the stream, the author devoted especial 

 attention to the conditions found there. At the time of the first 

 visit on July 21 the stream was ligh and, though it did not present 

 absolute maximum conditions, it gave a clear general idea of the 

 difficulties which the salmon have to meet during the periods of high 

 water. Upon returning in September the water had subsided very 

 distinctly and it was easy to compare the conditions then existing 

 with those seen on the former visit. Furthermore, visits were made 

 to the canyon on several occasions during the first week of study of 

 the Copper River. In that time the varying weather conditions 

 affected noticeably the height of the water. Taking all things into 

 account it seems certain that the conditions were not greatly modified 

 by changes in water level and that the salmon that ascend the stream 

 are subject to similar if not identical difficulties whether they go up 

 at high water or low. 

 35286°— 21 32 



