In the course of certain applications of Anopheline control measures by military malaria units in East Africa, including the use of residual DDT and airspraying with DDT, many of the gaps in our knowledge of Anopheline behaviour became only too apparent. The limitations placed upon the intelligent application of these measures by our lack of knowledge justified the investigations here recorded. At the same time the nature of the observations explains their apparent disconnectedness as well as their incompleteness. Time of Entry into Huts and Tents.The mosquitos in several small mud huts and bush tents near Taveta in Kenya were collected at three-hour intervals during the night on 51 occasions between November, 1944, and June, 1945, with the following results. (See also fig. 1.) Fig. 1.-Times of entry of mcsquitos into huts and tents.

Earlier papers in this series have described attempts to control tsetse flies by applying insecticides from aircraft. It was concluded that coarse aerosols, with mass median diameters of approximately 100 microns, were more effective than coarse sprays with mass median diameters of 350-700 microns, but that it would not be economically possible to control Glossina palpalis (R.-D.), the species associated with dense riverine vegetation, by applications from aircraft. Very promising results were obtained (Hocking & others, 1953) when coarse aerosols were applied to savannah woodland containing G. morsitans Westw. and G. swynnertoni Aust. In one experiment, where eight applications, each at a nominal dosage of 0-25 lb. of technical DDT (80 per cent. p,p'isomer) per acre, were made at intervals of two weeks, it is possible that G. morsitans, and even G. swynnertoni, would have been eradicated if the treated woodland had been isolated effectively enough to prevent immigration of flies. In another area, where a similar series of applications was made with BHC, the nominal dosage for each application being 0-25 lb. of crude BHC (12 per cent, y isomer) per acre, the reductions of fly populations were less spectacular, but these less favourable results were not attributed to a lesser effectiveness of the BHC, but to a combination of circumstances that led to a less effective series of applications.Several factors were considered to have reduced the effectiveness of the applications in the two experiments, quite apart from immigration which removed any possibility of eradication in the area treated with DDT. The aerosols were produced by allowing the insecticidal solutions to flow, under gravity, into modified exhaust systems of the engines of the aircraft; approximately 20 per cent, of the insecticide was thermally decomposed, and the emission rate fell considerably as the storage tanks emptied, so that wide variations in nominal dosages occurred during each sortie. It is also likely that much of the insecticide was contained in droplets that were too large to penetrate effectively into the resting places of the tsetse flies. Furthermore, the aerosols were not applied during the most suitable meteorological conditions of high temperature inversion and low wind speed (Yeo & Thompson, 1953, 1954, with the result that concentrations of aerosol near the ground were reduced. In the area treated with BHC an incomplete knowledge of the distribution of flies within the treated woodland, and a narrower overlap between successive sorties upon adjacent areas, also reduced the effectiveness of the applications.It was felt that the eradication of tsetse flies from a savannah habitat was possible, and could be achieved without any increase in the nominal dosage or the number of applications, if sufficient attention was paid to the factors mentioned above, and if there was no possibility of immigration of flies into the treated woodland. A new installation was therefore fitted to the aircraft, by which the aerosol was produced by emi...

A dieldrin emulsion spray containing 3 per cent, dieldrin (w/v) was applied once to the putative resting places of Glossina morsitans Westw. throughout a comparatively isolated 17-sq.-mile fly-belt in the Kabiganda Valley in southwestern Uganda between October 1957 and September 1958. Concentration areas of the fly consisted of one or more tall trees with associated understorey and thicket. The lower sides of the branches of an average of 6·75 such trees per acre were treated at a rate of about one-fifth of a pound of dieldrin per acre and at a cost of about £250 per sq. mile.Chemical analysis showed a deposit of approximately 0·8g. dieldrin per sq. metre on the surface sprayed, and although 90 per cent, of this had disappeared from the surface after about four months, it is thought that the application remains effective for this period.A very great reduction in the tsetse population was achieved. Small numbers of flies continued to be caught, but it was thought possible that these were being brought in from neighbouring valleys by buffalo, and, hence, that insecticidal application of this sort, if carried out over a wide enough area, might eradicate a tsetse population.

Summary An apparatus and method for testing the effect of atomized sprays on flies and mosquitoes are described. The apparatus consists of a revolving wire gauze cage placed in a thermostatically controlled chamber, the whole of which may be easily cleaned and freed from toxic residues. The insecticide is sprayed into the chamber by means of an Aerograph MP gun and distributed by means of a slow‐moving fan. When the insecticide has been injected and an interval allowed for the initial rapid fall of concentration, the movement of the cage is stopped and the insects are introduced into it by means of a special tube and plunger. The time required for paralysis to take place is recorded. After a given interval the insects are removed from the gauze cage and kept to ascertain the mortality. A technique for sampling the concentration of insecticide in the air space is described. The insecticide carrier is coloured with a dye, Sudan III for petroleum oil bases and methylene blue for water bases. The percentage of atomized material remaining in the atmosphere at any given time is determined by aspirating a known quantity of the atmosphere of the chamber through a sintered glass filter. The dye is retained in the filter, it is washed out with a measured quantity of liquid and compared with known standards. Reasons are given for the assumption that the dye molecules will behave in the same way as the insecticide molecules.This sampling method has been used to study the behaviour of a light oil, a heavy white oil and water at different degrees of atomization. Tables and graphs are given which show that, except with a fine atomization where most of the insecticide remains suspended for a considerable time, there is an initial rapid fall, which varies in amount with the degree of atomization. This initial fall is followed by a much more gradual decrease of concentration. The experiments show clearly that oil bases of different physical properties cannot be compared adequately unless a sampling method is used to ascertain the quantities of material remaining suspended. Where water bases are used it is shown that the time concentration curve varies not only with the degree of atomization, but also with the degree of saturation of the atmosphere before spraying. The concentration remains higher in an unsaturated atmosphere than in a saturated one.

Two experiments are described where applications of coarse aerosols have been made to areas of savannah woodland infested with tsetse flies (Glossina spp.).The applications were made at nominal dosages of 0·25 gallons per acre, which was equivalent to either 0·20 lb. of p.p′DDT per acre, or 0·03 lb. of γ BHC per acre. The coarse aerosols had mass median diameters of approximately 60 microns.In one experiment, carried out at Urambo, Tanganyika, a reduction of 95 per cent, was obtained in populations of G. morsitans Westw. This kill was somewhat lower than in many other experiments, a fact that can be attributed mainly to our inability to maintain the cycle of applications. Immigration of flies into the treated area caused a relatively rapid increase in fly numbers to levels comparable to the pre-treatment populations, and in this respect the experiment was a failure.The other experiment, in Lango County, Uganda, was highly successful, and reduced a population of G. morsitans to 0·05 per cent, of its pre-treatment level, and eradicated a small population of G. pallidipes Aust. It is indeed likely that no stable population now exists in the area, and that the very few flies caught there since the end of the applications have been wanderers from other infested woodland. The continued success of the experiment is considered to be due to the effective isolation of the area.Some brief comments are made upon the costs of the method, and on its value under conditions of land development in Africa.