floats unaccounted for went aground on the Oregon shore, which was 

 not checked because of engine trouble in the boat. Two days later 

 three white-top floats were found beached on the Washington shore 

 Just above Ellsworth cannery, but no floats were observed on the 

 Oregon shore. 



Observations of Dead Fish . Reports at various times had reached 

 the U. S. Fish and Wildlife Service and the U. S. Corps of Engineers 

 that great numbers of dead fish were floating below Bonneville Dam in 

 the vicinitj'' of Ellsworth cannery. We decided, therefore, to station 

 observers at Ellsworth cannery on the day following the peak of the 

 fall run of chinook salmon at Bonneville Dejn (Appendix, Table 2), on 

 the assumption that dead fish would appear in the river in greatest 

 numbers during and following the time when the fish were most numerous 

 at the dam. It had been determined from our recently completed float 

 experiments that dead fish originating at Bonneville would drift 

 dovmstream at about one and one-half miles per hoiir (Table 4), so 

 that dead fish from the dam shoiild cover the 33 miles to Ellsworth 

 cannery in about 43 hours, 



Salmon and trout sink immediately after death. Decomposition 

 soon begins, causing gas to form within the body of the fish. When a 

 sufficient quantity of gas has been formed, the carcass is buoyed up 

 and floats, provided, of course, that the gas is trapped within the 

 body of the fish, l/ 



Dead Fish Below Bonneville . The Ellsvxorth cannery wharf was 

 chosen for the place of observation because the main river channel 

 touches the end of the wharf, and it is in this main channel that 

 nearly all of the drift travels o Furthemiore, the wharf is high 

 enough above the water level that observers could readily see any 

 floating fish that might pass. Tvro men were stationed at this point, 

 one to watch for the floating fish and the other, in a power boat, 

 to pick up the fish. The number of dead fish observed at this point 

 is recorded in Table 5 and it can be seen that during 64 hovars of 

 observation covering a period of 8 days only 10 dead chinook salmon 

 were observed. On September 10, the observers examined both shores 

 by boat from Ellsworth, to Vfashougal, Washington; during seven hours 

 of observations, no dead salmon or steelhead trout were observed. On 

 September 19 the boat proceeded to Bonneville Dam and during the seven 

 and one-half hours of travel time, three dead chinook salmon and one 

 dead steelhead trout were observed. 



TT" Observations made by Charles B. wSde, Aquatic Biologist, U. Si 

 Corps of Engineers. The full text of his report is included 

 in Appendix II. 



16 



