FISHES FOR CONTROL OF MOSQUITOES. 21 



and in one specimen a large number of well-preserved chironomid 

 larvae and pupse. In 2 stomachs of fishes (No. 2356c, p. 16) taken at 

 the same time just outside of the pool were remains of 1 Culex larva, 

 a small dragonfly nymph, 2 small Planorbis, 1 mayfly nymph, a few* 

 filaments of Spirogyra, and fine debris containing remains of ento- 

 mostracans, protozoan tests, diatoms, and desmids. The larval count 

 had fallen to an average of 3.6 per dipper, just one-tenth of what it 

 was on the previous count, while in the check pool the number of young 

 larvse was so great that they could not be counted accurately, the 

 actual counts running to 40+ and 50-}- per dipper. 



On July 30 only one small mud minnow and no sunfish were found, 

 their absence probably being accounted for by the presence of a 

 water snake in the pool. The minnow (Table 2, No. 2361) contained 

 remains of Culex larvae, chironomids, undetermined insects, and some 

 small insect eggs. Mosquito counts ran from to 10 (average, about 

 4) per dipper, no pupae, and a few egg boats. In the check pool 

 breeding had increased if anything, the counts running from 45 to 

 over 80, of which about 20 per cent were large or full grown, the 

 rest of various sizes, largely newly hatched, pupae about 2 or 3 per 

 dipper and egg boats as in the opened pool. Mud minnows (No. 

 2362c) taken from the lake close to the pools contained 1 Aedes syl- 

 vestris larva, a few chironomid larvae and other insect remains, snails, 

 and a small quantity of algae and plant tissues. 



A third experiment (D 1) involving the mud minnow was on 

 a larger scale. An area of about one-third acre at the head of the 

 cove above Brooklyn Industrial Camp was much obstructed by drift- 

 wood (figs. 6, 7). Many logs had stranded on the bottom in shallow 

 water or had lodged on the projecting rocks and the few drowned 

 trees left standing. Eeinforced by large quantities of debris and 

 Lemna, together with some emergent and marginal vegetation, these 

 formed a barrier to the entrance of fishes. When this area was first 

 examined on July 5, culicine mosquito breeding was very general 

 but unequal, varying from an average of 30-[- larvae per sample in 

 the more obstructed parts near shore to 4 per sample in the outer 

 parts. Most of these were third and fourth stage larvae, together with 

 a few pupae of Culex pipiens and all ages of Aedes sylvestHs. On ac- 

 count of the many snags seining in this area was ineffectual, but 

 thorough dip-net explorations discovered no fishes except on the 

 outer margins. Tadpoles, newts, water bugs and beetles, many kinds 

 of insect larvae, etc., were taken in the abundant Ceratophyllum, It 

 is possible that a few mud minnows may have escaped detection, but 

 there were certainly not many in the more obstructed parts near 

 shore. 



On July 10 and 11 most of this area was cleared of obstructions (fig. 

 8). On the 11th large numbers of sunfishes and mud minnows were 

 seined on the border of the opened area. Stomachs of four of the 

 latter (Table 2, No. 2330, p. 17) yielded no mosquitoes but many 

 chironomid larA'^ae, some entomostracans, and considerable mud with 

 diatoms, desmids, etc. One contained some eggs of sunfishes which 

 were nesting a few yards away on a gravelly beach. 



Of fishes taken here on July 24 four mud minnows (Table 2, No. 

 2360d) were examined. The stomachs of two were empty, the others 

 contained many larvae easily recognized by their siphons as Culex 



