152 MOSQUITOES OF NORTH AMERICA 
campa appear to be restricted in habitat, each to a particular plant, where they 
prey upon the other mosquito inhabitants. 
Certain mosquito larve may be said to be partly cannibalistic, feeding upon 
their own kind and upon other species when pressed for food. This has been 
observed with Anopheles, and in Brazil Aédes calopus and Limatus durhamu 
have been shown to behave in the same way by Peryasst and his associates. 
It would seem, on the whole, that the presence of mosquito larve in water must 
be beneficial rather than the reverse, and this is probably the only good thing 
from the humanistic standpoint which can be said about mosquitoes. Of course 
there is no doubt as to the benefits derived from the larvee of Psorophora, Mega- 
rhinus, Lutzia and the like, since they destroy the larve of other biting mosqui- 
toes, including those forms which carry disease. As to the others, their effect 
upon the water on the whole must be purifying, and there is an idea in parts of 
Texas and others of the southern United States that where mosquito larvee occur 
in the rain-water tanks and wells—the tanks in many places furnishing the 
drinking water supply—this water is purer and safer than water which does not 
contain these larve. We have had drinking water served at table in small hotels 
in the South, which contained Culex or Aédes larve, and, on requesting a glass 
of water without larve in it, have been assured by the waiters that this was an 
evidence of the purity of the water. This possibly restricted popular idea may 
have a sound basis, since certainly mosquito larve eat many organisms and 
among these must often be such as are injurious to the human species. It must 
be remembered, however, that with bacilli this function may not in the least 
reduce their vitality and they may pass through the alimentary canal of the 
mosquito larva unharmed. 
Recently Boyce and Lewis have attempted to show that bacteria thrive better 
in the presence of mosquito larve, and this may be true under some circum- 
stances. They point out that there is some opposition to the destruction of the 
larvee of the yellow-fever mosquito (Aédes calopus) in receptacles with drinking 
water on account of their supposed purifying function. To test the truth of this 
belief they made a series of experiments. These consisted in placing larve of 
Culex and of Culiseta annulatus “in a flask of non-sterilized drinking-water, 
and comparing from day to day the number of bacteria present in the water with 
the number present in a control flask to which no larvee had been added. The 
number of bacteria was estimated by plating 1 cc. of the water in gelatine and 
incubating at 21° C. for 72 hours.” In all but one of the four experiments the 
bacteria showed greater increase in the flasks with mosquito larva. This in- 
crease might easily have been due to the introduction of further bacteria together 
with the mosquito larve, although it finds its readiest explanation in the organic 
substances discharged by the larve. The experiments are inconclusive as they 
were carried out under artificial conditions and as, furthermore, it is impossible 
to avoid all sources of error. The fact that the larve of different species differ 
in feeding habits also must not be lost sight of as it will affect the bacteria. 
There is great diversity in the behavior of mosquito larvee. Some live almost 
wholly at the surface of the water, others at the bottom, while many rise and 
