8 V. S. BUREAU OF FISHERIES. 



none were found. The aim was to visit each of these at weekly or fort- 

 nightly intervals, but this was found to be practicable in the case of 

 the more accessible stations only, such as those near Philadelphia and 

 Media and in Palisades Interstate Park during the summer of 1919. 

 The outlying stations were examined at longer and more irregular 

 intervals. 



Usually on each Adsit collections of fishes and other organisms 

 were made with minnow seine, dip net, and by hand. So far as 

 possible these were determined and recorded along with other eco- 

 logical and physical data. Samples of the fishes taken were dropped 

 immediately into 4 per cent formaldehyde to stop digestion and 

 preserve the stomach contents for laboratory study. The density 

 of mosquito breeding was determined by taking a number of sam- 

 ples of water (usually 10 at each point) in a small dipper. These 

 were made as uniformly as possible of 3 to 4 fluid ounces. The 

 number of mosquito larvse, pupae, and egg boats in each were counted 

 and recorded and samples preserved for identification. While this 

 does not give an accurate determination of the number of mosquitoes 

 present it does give a fair basis for comparing the rate of breeding 

 at different stations or at the same station on different days or under 

 changed conditions. If adult mosquitoes were flying, samples of 

 these also were preserved. This procedure was repeated at each 

 station as often and as thoroughly as circumstances would permit 

 and comparisons made between the fishes taken and the contents 

 of their stomachs at mosquito-breeding and nonbreeding stations. 



Numerous experiments were tried by modifying conditions in such 

 manner that fishes were admitted to places from which they had 

 been absent previously or, on the contrary, were debarred from 

 places to which they had had free access. In general, comparison 

 being made between stations in the same pond, the former were mos- 

 quito-breeding areas, the latter mosquito-free or nearly so.* These 

 two sets of experiments served as reciprocal checks on each other 

 and on the natural conditions remaining. 



Admission of the fishes was effected by cutting or otherwise re- 

 moving vegetation, removing stranded logs or debris, taking out 

 rocks, cutting channels through bars or banks, etc., thus giving 

 them natural access to new waters. In some cases actual transplan- 

 tation of certain species to small ponds or detached pools was made 

 on a small scale. 



In order to bar fishes from a closed area, an opposite procedure 

 was adopted. Most frequently wire screen was used for this pur- 

 pose, and while there was no uniform size or form of inclosure 

 a standard pen used in a number of cases measured 8 by 4 feet, 

 divided into two parts each 4 by 4 feet, one being* made of one-half 

 inch or three-eighths inch cellar window screen, the other of wire 

 mosquito screen, supported on a light wooden frame and held by 

 stakes driven into the bottom. The fine screen stopped all fishes, 

 even the smallest fry; the coarser one permitted the passage of 

 minnows up to about 2 inches long and sunfishes up to 1^ inches 



* In no case was a body of watpr selected for experiment in whicii mosquito breeding 

 had not been detected at some point or points within or in immediate proximity thereto, 

 as in a communicatinK pool or bordering marsli. A number of bodies of water apparently 

 suitable for breedinj; purposes were found from which moscjuito larvae were nearly or quite 

 absent throuffh the operation of causes other than the presence of fishes. These may be 

 discussed at another time. 



