22 U. S. BUREAU OF FISHERIES. 



pipiens. These made up the bulk of the food, the only other contents 

 being Lemna plants and one midge larva. Culex larvae at this point 

 ran regularly 2 to 4 per sample, except in some of the inaccessible 

 'pockets where the number rose to 20-f- to 30-|-. These were of all 

 sizes, including some of Aedes sylvestris and a few pupae. 



Of mud minnnows taken on July 30 (Table 2, No. 2364a, p. 17) 10 

 stomachs were studied ; 4 were empty, 5 others contained a total of 

 11+ Culex larvse and remains of a large number of chironomid lar- 

 vae in tubes, some fragments of other insects, some entomostracans, 3 

 Planorbis, and some protozoans and desmids. On this date general 

 mosquito breeding in these opened waters was practically abolished, 

 only an occasional larva being found except in some better protected 

 indentations of the shore where the counts averaged from 6 to 8. 

 Most of these were obstructed by rafts of Lemna and were doubtless 

 the source of most of the mosquito larvae found in the stomachs. 



These three experiments are typical of a number and are selected 

 because the results appear to be unequivocal. No change whatever 

 was effected in these natural breeding places of mosquitoes other than 

 to open them to the active life of the lake and especially to the small 

 fishes. Just as far as the latter penetrated, the density of the mos- 

 quito breeding steadily declined during July and August while ob- 

 servations continued. In undisturbed portions of each station left 

 as checks the original rate of breeding was maintained. Other sta- 

 tions at which similar experiments were conducted were sometimes 

 tampered with or may have been subject to other influences which 

 might contribute toward repression of mosquito breeding. The most 

 important of these are wind and wave action, from which these three 

 experiments were perfectly sheltered, the use of oil, which was not 

 employed in the vicinity of any of these stations, and the effects of 

 copper sulphate used in the lake against water bloom but not used 

 in this area during the progress of the experiments. 



The results, taken in connection with others, appear to the author to 

 establish the value of the mud minnow as a factor in mosquito re- 

 pression. It is, of course, not meant to credit this species with all or 

 even most of the reduction in numbers of larvae. As will be shown 

 later, that credit belongs in larger part to the common sunfish. 



The reverse experiment of excluding mud minnows as well as other 

 fishes from open areas free from mosquito larvae was made both near 

 Philadelphia and in Palisades Park by means of wire screens and 

 earth dams. In due time most of these pens became prolific breeding 

 places of culicine mosquitoes which grew to maturity undisturbed 

 except when the pens were opened, when the larvae quickly disap- 

 peared. This was repeated many times. As all fishes were excluded 

 it is not possible to differentiate the specific part played by Umbra. 

 There was not time to arrange experiments to attempt this differen- 

 tiation, but this may be done later. 



Prof. Smith (1904, p. 112) writes of the mud minnow, after stating 

 that one left in a bottle over night with a lot of wrigglers ate them 

 all, that — 



in these very pools known to be inhabited by these minnows there are always 

 plenty of larvae to be found, so that Mr. Seal is probably quite right in his con- 

 clusion that the species is unworthy of consideration in this connection. 



In a later paper (1910) Seal considerably softens this verdict. 



