230 BULLETIN OF THE BUREAU OF FISHERIES. 
By following the charted records—say, of the largemouth black bass, for instance— 
it is very instructive to note that through June and July, while both nymphs and fish 
were quite small, the former do not appear in the diet of the latter. The bass began to 
eat nymphs about the first of August, and during that month in 10 out of the 30 bass 
examined odonate food reached from 90 to 100 per cent. The fish continued to eat 
nymphs in large quantities through the fall months. The largest bass was examined 
September 2 and was 105 mm. in length; 95 per cent of its food consisted of dragonfly 
nymphs. During the autumn of the first year, therefore, the odonates supply a very 
large percentage of the food of the young fish, and many fish feed entirely upon them. 
Not all fish, however, will take odonate food, no matter how abundant it may be. 
The two examinations of buffalofish have been included in the table in order to show 
this fact. The first lot contained fish between 33 and 78 mm. in length, taken from 
pond 7D and several other ponds not in series D. These fish were all of just the right 
size to eat freely of the nymphs, judging by the records of the other fish enumerated. 
But with the exception of one which had eaten a few odonate eggs, probably acci- 
dentally, there was an entire absence of odonate food in their diet. Similarly, the 
350 larger buffalofish did not show a single instance of the presence of odonate food. 
We may reasonably conclude, therefore, that such food is not palatable to them and 
will be of practically no assistance in their culture. It is worthy of note in this con- 
nection, however, that Prof. Forbes found both dragonfly and damselfly nymphs in 
the stomachs of adult buffalofish. 
EvIDENCE FROM Exuvia.—There are also other methods of obtaining evidence 
besides an examination of the fish’s stomach. Ponds 1, 2, 3, and 4 are as nearly alike 
as possible in their conditions and environment, but differ considerably in their fish 
contents. Pond 1 was stocked in the spring of 1916 with about 4,500 small bass and 
bluegills, while ponds 2 and 3 contained only a comparatively small number of adult bass 
and bluegills, together with the season’s hatch of young, the latter, of course, being too 
small to eat any nymphs before‘autumn. There was no appreciable difference in the 
number of dragonflies hawking and ovipositing around these ponds, but there was a 
very marked difference in the number of nymph skins obtained along the margins of the 
ponds. The north shores of ponds 1 and 2 are of the same length and have a border of 
Carex stricta of the same width, but while this border yielded 450 nymph skins on 
pond 2, only 150 could be found on pond 1. ‘The difference in damselfly skins was even 
more marked; only 6 were found on the whole margin of pond 1, while a single reed 
stem in pond 2 yielded 21 skins, and the entire margin yielded a trifle over 500 skins. 
It would suggest, at least, that the young fish in pond 1 had reduced the numbers of 
nymphs by eating them, since they were of just the size to eat them freely. 
The same fact is even more markedly shown in the case of the Anax nymphs; a 
few were found in all of the ponds, but when pond 4, containing only adult buffalofish 
was drawn 200 nymphs were obtained and after the pond was refilled 100 more trans- 
formed and left their skins around the margin of the pond. Such disparity in numbers 
could hardly be due to a discriminating choice on the part of the adults as to the pond 
in which they laid their eggs. 
EVIDENCE FROM FEEDING NymMPHS ARTIFICIALLY TO FisH.—Best of all, the eating 
of nymphs by adult fish has been demonstrated by feeding the nymphs directly to 
them. Pond 6 is the smallest of all the ponds, but it contains the largest aumber of 
