8 FISHES IN RELATION TO MOSQUITO CONTROL. 
partly of a dam and partly of wire netting, was built across the 
pond. About 3,000 adult female Gambusia were placed in the larger 
compartment. No record was kept of the number of males which 
were introduced, as this is relatively unimportant,? but there were 
probably not more than 10 to each 100 females. The purpose of 
dividing the pond into two compartments, using a one-fourth inch 
wire screen for a portion of the pee was to give the young fish 
an opportunity to migrate to that section of the pond where they 
could not be followed by the adults. This was thought advantageous 
because of the cannibalistic habit of Gambusia, which the mother, 
in confinement at least, so impressively displays ae eating her own 
young oftentimes nearly as rapidly as they are born. The one- 
fourth inch mesh, however, proved to be a little too large, and some 
of the adults succeeded in getting through it. The reproduction in 
this hatchery, nevertheless, exceeded all expectations. This pond, 
being conveniently situated, furnished a ready supply of top minnows 
for aquaria, fountains, pools, and wells wherever they were found to 
be needed or as requests for them were received. 
A very useful net for collecting top minnows is a small bobbinet 
seine. ‘The one used by the writer was about 12 feet long and 3 feet 
deep. Such a net, if made of a good grade of netting, is light and 
durable, and it can be quickly and easily handled. A dipnet, also 
made of bobbinet, was used to some advantage in places where there 
was so much vegetation that a seine could not be operated. 
OBSERVATIONS AND EXPERIMENTS. 
Field observations were made at all ponds in the protected zone 
at more or less definite intervals of one week each throughout the 
investigation. These were often extended beyond the protected 
area for the purpose of obtaining checks on the effectiveness of the 
work within the zone. 
As early as March 29 a certain pond was found to be fairly alive 
with mosquito larve and pupe. Further investigation proved that 
no fish were present. Top minnows were then placed in the pond 
for the purpose of observing whether or not the fish would destroy 
the large numbers of immature mosquitoes. The fish, however, 
all died in less than 45 minutes. Since the pond was near that por- 
tion of the Georgia Chemical Works where sulphuric acid is manu- 
factured, the presence of a chemical fatal to fish life was at once 
suspected. A litmus-paper test gave a strongly acid reaction. 
While the first purpose of the experiment failed, it, nevertheless, 
was learned that mosquitoes can breed in water so strongly acid 
that Gambusia is killed almost instantly thereby. There, however, 
was another pond very near the acid pond and in appearance ver 
similar to it. This one was well supplied with top minnows in ad- 
dition to a few food fishes, and it was entirely free of mosquito 
larve. As it did not seem reasonable that mosquitoes would select the 
acid pond in preference to the unpolluted one for breeding purposes, 
the only sare logical conclusion was that the absence of mosquito 
larve in the latter was due to the presence of natural enemies or fish. 
Another situation very similar to the one discussed in the preced- 
ing paragraph was not far away, for there were nearly end to end an 
@ Hildebrand, Samuel F., Report, U. S. Commissioner of Fisheries, Appendix VI, 1917, p. 7. 
