FISHES FOR CONTROL OF MOSQUITOES. 31 



of mosquito breeding. It had recently been drained and refilled, 

 and at the time of the examination no larvse were found, though 

 the author had been assured that they were always present in con- 

 siderable numbers. The shores of the pond were heavily grown 

 witli brush to the water's edge, and the growth of emergents at 

 the head was dense. At the foot, toward the dam, the growth of 

 Elodea was very dense and the water was full of accumulated decay- 

 ing vegetation evolving decomposition gases. No fishes were seen 

 anywhere in the pond, and the seine brought to light only a very 

 few young roach and killifishes. Of four (Table 4, No. 189061, p. 29) 

 of the latter examined one contained a Culex larva, while the 

 same number of roach yielded no mosquito stages. In a pool be- 

 low the dam, however, the killies were present in considerable num- 

 bers, possibly having been carried out when the water was drawn 

 off. Of 10 (No. 18906g) of these examined the stomachs were mostly 

 empty. 



The facts that it is desired to emphasize are : ( 1 ) That the trans- 

 lucent killifish does eat mosquito larvse, to a limited extent at least ; 

 (2) that in the very pond where this species was most abundant and 

 the shore conditions most favorable to its action, mosquito breeding 

 was practically nonexistent; and (3) that in adjoining ponds where 

 fewer of the killies were present and where plants grew more luxuri- 

 antly, both Culex and Anopheles bred freely. ^¥liile the species was 

 met with in other ponds and creeks no observations of value were 

 made. Some experiments to test its practical value were suggested 

 to the Essex County commission, but nothing seems to have been 

 done, and there was no opportunity for personal experimentation. 



Little is known of the translucent killifish in relation to practical 

 mosquito control. Seal (?Vi Smith, 1904) advises against its introduc- 

 tion for this purpose " because it would destroy the eggs and young 

 of more valuable species, which are by nature better adapted to land- 

 locked or stagnant waters." Pearse's investigations show only 0.8 

 per cent of such food, and it is doubtful if this killy is greatly more 

 culpable in this respect than most of the fresh- water minnows. It is 

 certainly quite as well adapted to many mosquito-producing waters 

 and more efficacious as a mosquito destroyer than they. For muddy 

 ponds and sluggish streams the author would recommend it much in 

 preference to the common killy, but experimentation is much to be 

 desired. 



TOP MINNOW (Gambusia affinis). 



The natural range of this top minnow is from Delaware through- 

 out the South Atlantic and Gulf States and up the Mississippi Val- 

 ley to Illinois. It is an inhabitant of shallow estuaries, sluggish 

 creeks, lagoons, ponds, and marshes and is equally at home in 

 lirackish and fresh waters. While preferring clean water, it will 

 live in that which is most extremely foul, as the writer observed in 

 drying borrow pits in Louisiana. It is excessively abundant in the 

 South. 



The value of this little fish in the natural suppression of mos- 

 quitoes, and especially of the anopheline mosquitoes, has long been 

 recognized, and in the South, during the war period and since, it has 

 been extensively and successfully utilized in the antimalaria cam- 



87028°— 22 3 



