FISHES FOR CONTROL OF MOSQUITOES. 35 



was easily overcome by making a screened overflow a couple of inches 

 below the top of the barrel. 



The above account shows that Gambiisia may be successfully em- 

 ployed north of its natural range in controlling mosquitoes in water 

 gardens and other small j)onds heavily grown to vegetation. A few 

 specimens introduced in the early spring will have increased to 

 effective numbers by the time mosquito and especially Anopheles 

 breeding becomes active. The water temperature of 60° F. for plant- 

 ing was selected arbitrarily, and it may be that they could be planted 

 puccessfully at lower temperatures, but it is safer to avoid too great 

 a shock. The greater number of individuals should be females, but 

 it is desirable to have some males, though as the species is viviparous 

 it may be that the first brood from pregnant females would supply 

 this need. Of course the larger tlie number introduced the sooner 

 will effective control be established, but under favorable conditions 

 they increase with astonishing rapidity. It is also important in 

 these small ponds that i:)redacious fishes and other enemies should 

 be removed as thoroughly as possible. The adults also devour the 

 young, especially when other food is scarce. For this reason the pop- 

 ulation in a pond will in time become self-limiting. To postpone 

 this as long as possible it woidd be desirable at the beginning of the 

 season either to isolate the breeding fishes in a pen of one-fourth inch 

 cellar window screen through which the young may escape or to 

 provide retreats for the young. As maximum destruction of mos- 

 (juitoes depends upon the hunger of the fishes they should not be 

 artificially fed. 



After a supply of Gambusia is once secured it is a simple matter 

 to maintain it by removing in the fall to a greenhouse or other suit- 

 able place indoors a sufficient number to provide brood fish for re- 

 stocking in the spring. They are easily caught with a dip net or 

 minnow seine, and several hundred may be kept in an ordinary tub 

 in which are a few spraj^s of Elodea or other water plant. They may 

 be fed with fish foods, yolks of hard-boiled eggs, or boiled liver 

 ground fine, or still better with mosquito larvse or entomostracans. 

 As hundreds of Anopheles-breeding ornamental ponds are found in 

 country places, parks, and the suburbs of every large city aquarium 

 dealers might find a profitable business in furnishing Gambusia for 

 stocking them. 



Another type of mosquito-breeding habitat, where Gambusia 

 would undoubtedly prove extremely valuable, is found in the very 

 shallow plant-choked waters about the heads of many ponds and in 

 swampy areas generally. But such areas are often too large for an- 

 nual stocking. The planting of a permanent self-perpetuating stock 

 only would suffice. Up to the present such stocking in the latitude 

 of Philadelphia has failed. Just why Gambusia will not live over 

 winter is not clear. The winter of 1920-21 was no more severe than 

 those to which the top minnow is frequently exposed within its nat- 

 uial range, yet in four ponds in which they thrived during the pre- 

 ceding summer not a single one appears to have survived. In an 

 attempt to solve this second problem it is proposed to search the 

 northern border of this fish's range for a cold-resistant race or to 

 attempt the isolation of such a race by selection and possibly by 

 hybridization. 



