Mosquito Remedies and Enemies 131 



have found Culex and Anopheles larvae ten times in com- 

 pany with fish ; while Mr. Aitken reports that he never 

 found larvae and fish in the same pool, adding that of all 

 larvicides the most effectual in the case of Anopheles are 

 little fishes. These discrepancies in no wise discredit 

 the accuracy of the observations they record, but clearly 

 show that the tastes of different species of fish differ like 

 those of other orders of animals. Impressed with the 

 economic importance of fish as factors in the destruction 

 of gnats, a number of observations have been made by 

 us with a view of determining the larva-feeding capacity 

 of the various species that inhabit our local waters, from 

 which we have reached the conclusion that for the maxi- 

 mum of efficiency no species so nearly suffices as does 

 top minnows {Gambusia affinis). It is true that they are 

 omnivorous and soft-rayed (objections urged against their 

 general employment as mosquito exterminators, and in 

 favour of wholly carnivorous and spine-rayed species), but 

 we are persuaded that their diminutive size, extreme vital- 

 ity, restless activity, exceptional reproductive capability, 

 and surface feeding habit will more than counterbalance 

 the advantages claimed for those species of wholly carniv- 

 orous appetite and with greater freedom from molestation 

 bylargerfishes. Infactwe have long since learnedtheutter 

 futility of searching for culicid larvae in pools that harbour 

 top minnows. The presence of minnows is to us sufficient 

 guarantee of the absence of mosquito larvae. Observa- 

 tions during the summer of 1902 fully justify the latter 

 statement. In a low-dying wooded area near a public road 

 not far from the city of Baton Rouge there exist two de- 

 pressions, one in either bed of the branches of a small 

 stream that divides a short distance above the depressions. 

 These depressions are similar in contour formation and 



