American Fisheries Society. 125 
weighed, and it was found that the heaviest fish were those that 
had the shortest fasts, they were averaging 35 and 37 grains re- 
spectively ; but the smallest fish were not much smaller, 28 grains 
for the 15-day fast of the first series and 29 grains for the 20-day 
fasts of both series. Now comparing these with a control lot 
consisting of 2,986 fish that had been fed promptly and continu- 
ously, we find that the latter had not attained any greater size 
than the smallest of the first series of fasters and only a single 
grain larger than the smallest of the second series, and that out 
of the 8 lots of fasters there were 5 lots that actually outstripped 
in growth the full-fed fish. 
Turning now to.the brook trout, we find that in August and 
September the losses were very much heavier than with the At- 
lantie salmon, but the distribution of the losses was such as to 
forbid any decided conclusions favorable or unfavorable to the 
fasts. October 16th they were all weighed, with the very inter- 
esting result that the larger fish were among those that had fast- 
ed the longest, the average being, from the short fasts upward, 99 
grains, 102 grains, 109 grains and 127 grains, successively, and 
the last named, which had fasted 19 days, had attained an aver- 
age weight 14 grains (over 10 per cent.) greater than that of a 
lot of 3,595 trout that had not fasted at all. 
The lake trout suffered so lightly during the late summer and 
autumn as to hardly afford any data except of a negative char- 
acter (indicating that the fasting had done no harm) and in 
October the weighing showed that the two lots of greatest mean 
weight were those that had fasted the longest. 
The mortality among the silver salmon was too light between 
June and October to afford data for comparisons, and the Octo- 
ber weighing afforded data for only the negative conclusions that, 
when compared with each other, the long fasts had been little if 
any more detrimental than the brief ones; but when compared 
with a control lot of 2,441 salmon of the same species and age, 
it is found that these early and full-fed fish were decidely heavier 
than any of the fasters. 
I must call your attention to certain points in which we are 
hable to err in our interpretation of the results above stated. 
Let us, in illustration, take up again the observations made on 
the brook trout. It was observed that the survivors of the brook 
