NATURAL HISTORY OF THE QUINNAT SALMON. 89 
the Battle Creek fishery, with the wings extending obliquely upstream, their ends 
being about 10 feet apart. The fish were deflected by the wings to the middle por- 
tion of the net, and found their way through the funnel into the bag. No effective 
means could be devised to prevent the funnel from becoming choked with leaves or 
other trash, which often happened within an hour or two after the net was set. 
There were many adult salmon below the rack, and they often tore the net with their 
teeth and frequently got fast in the funnel. Part of the time the net was set during 
the day, more often during the night. Sometimes it was set for only an hour or two 
during the night. The following is a record of the catch, showing the date, the time 
of day, and the number and size of the fry taken: 
Record of salmon fry taken in trap at Battle Creek fishery , Oct. 7 to Nov. 30, 1898. 
X ote. — The numbers in the column headed 1.5 ± and 5. ± indicate the number taken that were about 
1.5 inches long, or about 5 inches long, as the case may be, but were released without measuring. 
From the preceding record it will be seen that all of the fry (not including the 
parrs) were practically of the same size, 1.5 inches long. Of the 322 fry examined, 
only two were over 1.6 inches long, one being 1.7, the other 1.8. The 1.5-inch 
specimens had just absorbed the yolk-sac. Indeed, there was often a small amount 
of yolk remaining in the body, although the sac had disappeared. The size of these 
specimens shows that they begin their downstream migration as soon as they begin 
swimming, or at the age of six weeks; their continuing the same size during the 
two months shows that practically all start downstream at the same age. If part 
of them had held back for two or three weeks, this would have been indicated by 
a greater variation in size. 
