American Fisheries Society. 85 
the future lead to similar results as these that I have reported, 
then we should feel quite safe to do such things. 
Mr. Clark: J think Mr. Atkins has stated in his discussion 
that in 1904 the fish became mixed. 
A. Yes. 
Mr. Clark: That was the first time we heard you say any- 
thing about 1904. In the paper you did not give us those 
figures. 
Mr. Atkins: I mentioned the fact that I tried it in 1904. 
Q. Did you find the percentages run the same in 1904 as 
in 1905? 
A. Yes, about the same. 
Mr. Marsh: Was every remnant of the sac in the 19 day lot 
absorbed before the period of fasting began ? 
A. Yes. 
Mr. Dean: This last spring we took a lot of 1,000 trout up 
to the spring to test the water; it was over 2 miles up there, 
and we could not feed the fish more than once a day; but they 
did as well or better than those fish that had been left in the 
hatchery. But in taking out fish for experiments, oftentimes 
we take them out of a lot of fish, and necessarily dip out the best 
ones; and you leave the inferior fish in the trough for control; 
as far as percentage went, those that went to the spring did 
better than those that were left in the trough. When there are 
a lot of fish in a trough, the weaker ones go to the lower end, 
and you usually go to the head of the trough to dip out fish for 
experiments, and thus get a superior lot. 
Mr. Atkins: [| thought of that point after the experiments 
had been concluded, and I asked my foreman, who had direct 
charge of the matter, whether he tried to select the best fish, 
or whether he tried to select just a fair average of them, and 
he told me that he tried to avoid selecting the best fish but to 
make a fair average. His effort was in that direction; but of 
course it is quite possible that unintentionally the men who 
