MOSQUITO LARViE AND THEIR NATURAL ENEMIES. 215 



through a nutnber of vessels containing sterilised water (to wash them free 

 from bacteria), are placed in this. They do not develop. If, however, 

 they are supplied with sterilised algae, they develop. These experiments 

 seem to show that the larva? are dependent on algae, or certain decayed 

 vegetable matter. These factors may be of some |.ractical importance, as I 

 will attem[)t to show later when discussing practical means of lessening tlu^ 

 larvae. I will now turn to the natural enemies of the larvae, and describe 

 them, together with experiments I have done to determine their practical 

 possibilities. 



" The chief among these in this island are the fish here known as ' Millions.' 

 I believe there are two varieties of these (I am forwarding specimens of the 

 second variety to the British Museum) : the one technically known a.s 

 GirarcUnus pwciloides exists in the swamp at the north end of the island, and 

 the other variety at the south end. The greater part of the island had none 

 of these fish ; I have recently distributed them over certain districts. 



'' My experiments have all been made with the former, and were made to 

 ascertain their power of destroying larvae and their adaptability to varying 

 surroundings. 



(1.) I obtained about twenty fish from the swamps of Grros-Islet and 

 placed them in two old spirit-barrels, which had previously been 

 soaked in tanks for two months. In these barrels I have now bred 

 many thousands, which have been variously distributed. 



(2.) At first the fish would not live in iron tanks ; but by mixing a little 

 water from rusty tanks with a large quantity of rain water, and 

 gradually increasing the percentage of the rusty tank water, I 

 obtained fish which would live and multiply in iron tanks. 



(3.) In the same way I succeeded in immunising fish to darkness. 



(4.) By gradually adding water from brackish swamps they are immunis 'd 

 to brackish water. 



(5.) I have had great difficulty in getting fish to live and multiply in water 

 at and above 100°Fahr. ; but I have now succeeded, and the difficulty 

 I believe was entirely due to the dark incubator with which I was 

 obliged to work, layers of fungi quickly appearing on the surface of 

 the water, and their food material quickly decom[;osing. Continually 

 changing the water and adding fresh food and allowing the sunlight 

 to play upon the water each day, has enabled me to get fish to flourish 

 at these higher temperatures. I have had difficulties with the lower 

 temperatures ; but the experiments conclusively show that these 

 fish can be gradually immunised to temperatures ranging from 

 101*5° Fahr. to 55"^ Fahr. ; and there appears to me no reason wliy 

 higher and lower temperatures should not be emploved, if a con- 

 siderable period of time is used to obtain the inimunitv. 



