418 JOURNAL, BOMBAY NATURAL HISTORY SOCIETY, Vol. XX. 



beautiful eyes which are large and bright yellow in colour. The 

 " Kazari " is a most active fish and very voracious, destroying 

 hundreds of mosquito larvaa in a very short time. It appears to be 

 a surface feeder and quickly swallows anopheles larvae and pupa3, 

 and I have seen it jump several inches out of the water after a flying 

 adult insect. Altogether the " Kazari " appears to be a most ex- 

 cellent fish to place in wells, in tanks and to keep in ornamental 

 fountains. The third fish Avhich Mr. Kinnear tells me is Polyacan- 

 thus cupanus was caught in the Gowalia tank. It is quite small, 

 about li to 2\ inches in length, with crimson underfills and a 

 peculiar whisp on its tail. It is not a very quick feeder and has 

 only a small mouth. It appears to eat mosquito pupae readily but 

 not nearly so quickly as the Kazari fish. The Chilwa (Ohihva 

 arqentia) which has sometimes been recommended for mosquito 

 destruction does not appear to be nearly as efficient as either of the 

 three fish I have just described. It is not a surface feeder and will 

 leave anopheles larvee untouched for long periods. The mud fish 

 common in tanks and bheels do not appear to feed readily upon 

 mosquito larvee. I obtained two species of fish of this kind from 

 the Gowalia tank but they would not eat either larvae or pupae. I 

 do not know the names of these fish but in Assam similar fish are 

 called " Singa " and " Mangur " respectively. " Jingas " or 

 fresh-water prawns ( called " Chingri " in Assam) will not touch 

 mosquito larvae. 



The presence of water-weed in a tank, well or cistern is often a 

 source of danger, forming as it does an excellent shelter for 

 mosquito larvae. Some water plants, notably duck-weed or Lemna 

 and Azolla have been suggested as being useful in preventing the 

 breeding of mosquitoes. These suggestions do not appear to have 

 been based upon careful observation and experiment. In my ex- 

 perience these plants are of little or no value in preventing the 

 presence of mosquito larvae. But there is a weed which I have met 

 with in Bombay and elsewhere which under certain circumstances 

 appears to be a useful preventive of mosquito larvae. This plant is 

 the rootless duck-weed or Wolffia arhiza. It takes the form of 

 small bright green round grains without stem, roots or leaves. 

 These grains float on the surface of the water just like a scum of 



