PRACTICAL USE OF NATURAL ENEMIES OF MOSQUITOES. 69 



the best insect feeders. The temperature of the water ranged from 

 74° to 87° F. Careful experiments were made with 10-gallon milk cans, 

 in order to determine the conditions under which the fish could be most 

 successfully transported to Hawaii. These experiments included 

 observations on temperature of the water and on changing the water, 

 and from these experiments was ascertained the necessary informa- 

 tion in regard to the frequency of changing and the fact that best 

 results could be obtained by transporting them in water of the normal 

 temperature. The three most abundant species, Gamhusia affinis, 

 Fundulus grandis, and Mollinesia, were then collected and about 

 75 were placed in each can, a 20-gallon tin tank full of water being 

 taken along as a supply reservoir. Mr. Scale left Seabrook on Septem- 

 ber 4, 1905. On the journey the fish were fed sparingly every morn- 

 ing at 8 o'clock on prepared fish food, finely ground liver, or hard- 

 boiled eggs. At half past nine one-half of the water in each can was 

 drawn off from the bottom, thus cleaning the cans and removing un- 

 eaten food and excrement. An equal amount of fresh water was 

 added. At noon the cans were aerated by means of a large bicycle 

 pump, a sponge being tied over the end of the hose to separate the air 

 into fine currents. At four in the afternoon 2 gallons of water were 

 drawn off from the bottom and 2 gallons of fresh water put in, and the 

 aeration was repeated just before bedtime. Careful tests of water at 

 each place of changing were made by experimenting with two fish. 

 At El Paso, Tex., there was so much alkali in the new water that the 

 fish were killed; at Los Angeles and at San Francisco the water was 

 good. Twelve fish died between Galveston and San Francisco, and 

 15 between San Francisco and Honolulu. Honolulu was reached 

 on September 15, 1905, with a loss of 27 out of approximately 450 

 fish. On arrival the fish were placed in the breeding ponds prepared 

 for them at Moanalua, near Honolulu, where four ponds had been 

 made ready. The fish thrived in all of the ponds almost equally well. 

 They were protected by screens from predatory fish and from being 

 carried out to sea by a freshet. In an official bulletin issued July 25, 

 1907, Mr. Van Dine reported that the fish had multiplied rapidly and 

 from the few hundred introduced several hundred thousand had been 

 bred and distributed. They had proved very effective against mos- 

 quito larvae and also against mosquito egg-masses. Later advices 

 show that the good work is continuing, and the experiment seems to 

 have been a great success. 



FISH IN THE WEST INDIES. 



Girardinus pceciloides, a small top-minnow, occurs very abundantly 

 in Barbados, where the popular name '^ millions" has been applied 

 to them. This fish is very small in size, the grown female measuring 

 about IJ inches in length, while the male is much smaller. The 



