FISHES IN RELATION TO MOSQUITO CONTROL. 9 
acid swamp and the pond previously referred to as the hatchery. 
Anopheles bred in the acid swamp throughout the season, except as 
interrupted from time to time by the application of oil. The hatch- 
ery pond, which was evidently well suited for the support of Anoph- 
eles larvee, was, nevertheless, free from them, except when the 
So became dense and provided protection. 
n April 1, mosquito larve of the Culex type were found in a ditch, 
in several pools, and in a large pond. ‘These waters were all free of 
débris and vegetation. The pond had steep shores, and it was very 
clean. It was one of the newer brickyard ponds, and no fish were 
present. All of these places, being close together, were stocked with 
Gambusia at the same time. Only 6 fish were placed in each of the 
pools, which were about 10 to 12 feet long, about 2 feet wide, and very 
shallow. Each pool supported thousands of mosquito larve, but 
in about two weeks they were made entirely free of wrigglers by the 
fish and remained so until they became dry later in the season. The 
resence of comparatively few skins showed that not many of the 
arve reached the adult stage. The ditch referred to was approxi- 
mately 30 feet long and 1 foot wide. Mosquito larve were especially 
abundant in it. Aibiout 200 top minnows were placed there, and in 
two weeks it was completely free of wrigglers. Mosquito larve 
were seen only along the shore of the big pond. It was at first stocked 
with about 1,500 Gambusia, but later several thousand more were 
added. The larve in this pond, too, disappeared in about two weeks 
from the time the first fish were introduced and none were again 
seen until September. By that time the shores had become over- 
grown with vegetation which furnished protection for Anopheles 
larve against fish. This vegetation was cut and the shores were 
raked. A large school of top minnows followed the workmen, de- 
stroying the immature mosquito and other insect larvee as quickly as 
their hiding places were destroyed. 
It was possible in several instances to connect ditches and swamps 
which were thickly infested with mosquito larvee with ponds that were 
well supplied with top minnows. Wherever this was done, large 
numbers of fish entered these waters and destroyed the mosquito 
larvee in a surprisingly short time. 
For the purpose of comparison and as further evidence of the value 
of Gambusia in controlling mosquito breeding two other small ponds 
are worthy of mention. ‘These ponds are situated at the intersection 
of the tracks of the Georgia Central and the Belt Line railroads, and 
they are of about equal size. The top minnow had reached one of 
the ponds from an unknown source, and it was entirely free of 
mosquito larve. The other was without fish, and mosquitoes were 
breeding in it in abundance. The inference, in the absence of any 
evidence to the contrary, of course, is that the top minnows destroyed 
the mosquito larve in the first pond. 
During the latter part of October two new brickyard ponds were 
found to be breeding large numbers of mosquito larve. The ponds 
had become supplied with some vegetation; the mosquito larve, 
however, were not confined to these hiding ea but were quite 
generally distributed over the ponds, and could be seen in perfectly 
clear water. Anopheles larve previously had been noticed only once 
away from all protection, and then, as now, in a pond not stocked with 
fish. On October 23 about 1,000 Gambusia were placed in one pond 
