348 
MALARIA 
marked differences in food preferences 
among species bnt sometimes by geographi¬ 
cal variations in the habits of a given spe¬ 
cies. Moreover, some vector species, like 
A. minimus var. flavirostris in the Philip¬ 
pines, feed on man or animal as occasion 
permits without a well-marked preference. 
Therefore, an experiment on zooprophy¬ 
laxis requires as much information as pos¬ 
sible about feeding habits of the local 
anopheline suspected of carrying malaria. 
Precipitin tests to determine the usual 
sources of blood meals of the vector are 
essential. For these the method of Rice and 
Barber (1935) is most suitable. It is also 
useful to make census studies of animal 
and human populations in relation to 
malarial prevalence, to study the flight, 
resting habits and time of feeding of the 
local vector. 
In such experiments some attention 
should be paid to spacing between the ani¬ 
mal barrier and the humans it is designed 
to protect for, when dealing with anoph- 
elines that have no strong preferences, it 
is possible to increase rather than decrease 
the danger by stabling animals too near a 
house (Russell 1934). In spite of a great 
deal of speculation and preliminary study, 
it is not possible to cite a definitely success¬ 
ful attempt to control malaria by zoopro¬ 
phylaxis. 
Naturalistic Measures Against Mosquito 
Larvae 
Drying by planting. Not much can be 
said for attempts to control malaria by 
planting trees to dry water-logged land, 
although occasional reports have appeared 
regarding the use of such trees as Euca¬ 
lyptus, Cassia, and Casuarina. Probably 
orthodox drainage would in most cases be 
preferable. However, Hopkins (1940) re¬ 
ports some success in using afforestation 
with Eucalyptus robusta to dry Uganda 
swamps where A. gambiae and A. funestus 
breed. 
Altering flora. One sometimes observes 
the disappearance of anopheline larvae fol¬ 
lowing some natural alteration in the flora 
of a breeding place. For example, a growth 
of Wolffia arrhiza over the surface of a 
pool will inhibit all mosquito breeding. 
This effect has also been reported in the 
presence of Lemna minor and of Salvinia 
cucullata and Azolla. It might be possible 
to use this natural force in mosquito con¬ 
trol, but at present too little is known 
about transplanting such plants and en¬ 
couraging the rate and luxuriance of their 
growth. 
Some success has ’ followed procedures 
for killing top-algae and stimulating bot¬ 
tom blue-green algae in salt-water fish 
ponds in Java. So, too, it is reported that 
by introducing a species of herbivorous 
fish, Puntus javanicus, into fresh-water 
fish ponds in Java the edges were kept 
clean-weeded and larva control by Panchax 
fishes was obtained (Hackett et al. 1938). 
However, one must avoid altering flora in 
such a manner that vector species are at¬ 
tracted rather than controlled by the 
changed conditions. 
Agitating water surface. Although in¬ 
termittent agitation in agricultural and 
domestic wells and gentle swells in a tidal 
basin are usually not sufficient to deter 
anopheline breeding, mosquito larvae and 
pupae are not found in water continuously 
ruffled by waves. Occasionally this method 
may be used in mosquito control. 
Scharff (1935) used agitation success¬ 
fully in some 200 agricultural pools on 
Penang Hill. He kept them free from A. 
maculatus breeding by changing the water 
supply line from small ditches to bamboo 
pipes with outlets four to six feet above 
the pools. 
Kibbey (1923, 1925) reported using 
launches to create wave action for the con¬ 
trol of A. quadrimaculatus breeding on some 
Alabama lakes. This device stranded float¬ 
age and, in conjunction with occasional 
hand removal of vegetation and debris, was 
effective. 
Experiments in Madras with wind-acti¬ 
vated “diddlers” were not successful in a 
small pool, but it appeared that further 
research might evolve practical control of 
this sort (Russell and Jacob i939a). 
Pollution. Anophelines in general will 
