32 U. S. BUREAU OF FISHERIES. 



paign. The conditions of its utilization have been especially worked 

 out by Hildebrand (1919), and some of its ecological relations by 

 Barney and Anson (1921). For the purpose of mosquito control 

 Gambusia is almost ideal. Its natural predilection for the aquatic 

 stages of mosquitoes as food, its vivaparity and fecundity, its hardi- 

 ness and adaptability to a great variety of aquatic habitats, and par- 

 ticularly its fitness to the marsh and swamp associations to which 

 mosquitoes naturally belong, its top-feeding habits, small size, great 

 activit}^, and proneness to penetrate into the shallowest waters and 

 dense vegetation, and its wide distribution and abundance are among 

 its outstanding merits. Within its natural geographical range its 

 value may be taken as fully established, and it has been successfully 

 jDlanted and employed against mosquitoes in other warm districts, as 

 in the Hawaiian and Philippine Islands. A related species has been 

 utilized recently in Mexico. 



As long ago as 1904 Heal experimented with this fish and pointed 

 out its great promise as supplementing the native fishes found in 

 New Jersey, especially in relation to the anophelines. After a pre- 

 vious thorough and futile search for Gambusia had been made in 

 southern New Jersey, Seal in 1905 planted 10,000 individuals. Ap- 

 parently this attempt at colonization failed unless the Gambusia 

 found by Fowler (1907) in small creeks emptying into Delaware Bay 

 in Cape May County in 1907 were derived therefrom, which seems 

 improbable. At the time of Fowler's discovery they were abundant in 

 small ditches and creeks in the vicinity of Goshen. In July, 1918, the 

 author visited this locality with Mr. Fowler in the hope of securing 

 a supply for experimentation. Not the slightest trace of the top 

 minnow could be found anywhere in the neighborhood, though Fun- 

 dulus was very abundant in the very creeks and ditches from which 

 the Gambusia had been taken. The previous winter had been a very 

 severe one, and it was learned from the natives that all of these 

 creeks had been frozen to the bottom. The resulting mortality among 

 many of the animals inhabiting these waters, notably the blue crabs, 

 was reported to have been very heavy. It was therefore concluded 

 that the Gambusia had been exterminated by the cold. It seems clear 

 that the established natural range of Gambusia does not extend north 

 of Delaware. Some inquiries were made to determine whether this 

 northern limit fluctuates with the degree of seA^erity of the winters, 

 but no reliable data were secured. 



(iambusia is so eminentl}^ fitted as a factor in the biological con- 

 trol of anopheline mosquitoes within its natural range that the de- 

 sirability of extending the area of its usefulness northward seems 

 unquestioned. The determination of a means of accomplishing this 

 is of some importance. A few experiments have therefore been tried. 

 On August 21, 1918, 200 Gambusia received through the Bureau of 

 Fisheries was introduced into a small ornamental pond (fig. 14) on 

 the campus of the University of Pennsylvania from which all other 

 fishes had been removed previously. Their increase Avas so rapid that 

 by the middle of September many young could be seen everywhere 

 about the edges of the pond which was kept nearly free from mos- 

 quito breeding for the remainder of the season. The following winter 

 was a very mild one, and there was a light covering of ice on the pond 

 foi- only two or three Aveeks. During the warm Aveather of Febru- 



