An investigation has been made into the distribution and bionomics of Ornithodoros moubata (Murr.) in East Africa in relation to the incidence of relapsing fever, and a survey has been made of the infestation in over 4,600 African huts, together with the temperature and relative humidity conditions.

In Kenya Colony, relapsing fever is endemic in the high rainfall areas of Meru, Nyeri and Taita Districts. These habitats are cool and wet with a mean microhabitat temperature of 71°F. and a relative humidity of 86 per cent. Tick infestations were relatively sparse and were rare in the hot and dry climate of Embu District, the base of the Taita Hills and generally over all such country in Kenya.

In Tanganyika Territory, relapsing fever is widespread, and the most striking difference was the relatively much greater abundance of the tick, especially in the dry central areas. It is pointed out that although relapsing fever is most prevalent in the north-west, endemicity is at a lower level than in Kenya, and decreases towards the south-east, indicating that the degree of incidence of the disease does not conform with that of the vector.

In the Digo District, south of Mombasa on the Kenya coast, ticks showed a reversal in their choice of microclimate from those in the cool highlands and were numerous in hot, moist conditions. The incidence of the disease was very low.

O. moubata was widespread in the Usambara Mountain area of Tanganyika. Ticks were most numerous in the cool, wet conditions above 4,000 ft., but were also abundant in the hot, moist foothills and plains, whereas they were absent in hot and dry country at the base of the Taita Hills in Kenya 80 miles to the north.

As humidity appeared to be a foremost factor affecting the distribution of O. moubata it was not possible to evaluate clearly the effects of temperature. It is suggested that all the conflicting evidence of the relationship of the tick populations to microclimate and the incidence of relapsing fever may be explained only by introducing a hypothesis of biological variation in the tick itself. It is shown that there are two peaks of greatest abundance, at relative humidities of 86 and 67 to 68 per cent. respectively, and it is suggested that these two peaks represent the distribution of two hypothetical hut-haunting biological forms.

An examination of the blood-meals from pooled catches by the precipitin test showed that in the cool and wet habitats of the Kenya highlands and the north-west of Tanganyika, 94 per cent. of the recognisable feeds were on man and only 2 per cent. on fowls. In the hot and moist habitats of Digo and the low-lying area between Digo and the Usambara Mountains, 18 per cent. were on man and 78 per cent. on fowl. In the mainly warm and moist habitats of the Usambara Mountains and the area bordering the south-east of Lake Victoria, 73 per cent. were on man and 22 per cent. on fowls.

It is therefore suggested that there are two biological forms of O. moubata found in huts, one feeding on man and the other feeding on fowls. The former is found in huts at high altitudes in areas having a cool and wet climate; it is essentially a human parasite showing a marked preference for the blood of man while ignoring the presence of fowls however numerous or available. It occurs in greatest abundance at a relative humidity of about 86 per cent. It is found at relatively low temperatures from 67° to 75°F. It is absent in areas where the microclimate is consistently over 90 per cent. R.H. and may not occur where it is consistently lower than about 74 per cent.

The form that feeds on fowls appears to possess a tolerance to a wide range of temperature and R.H., occurring in greatest abundance at 67 to 68 per cent. R.H. It is found at temperatures from 68° to 87°F. It is more resistant to starvation than the form that feeds on man.

Busseola fusca (Fuller) is a serious pest of maize and sorghum in Tanganyika where it occurs mainly at altitudes of 4,000 ft. and over.

Laboratory work on the life-cycle and on the diapause of the larva is described. Diapause is terminated by contact with water.

Field work at Nzega, Western Province, indicates that the first sowings, made within two to three weeks of the beginning of the rains, are likely to be heavily attacked by larvae of the first generation. These are the progeny of a “raininduced” flight of adults of Busseola derived from diapause larvae in the maize and sorghum left in the fields from the previous season, the larvae being induced to pupate when the stems are wetted by rain.

The number of generations in the main crop season, November–June/July, is two. A third generation occurs in sorghum tillers produced after June/July.

No alternate hosts of importance comparable with maize and sorghum have been found.

Cultural control by burning the crop residues left in the field after harvest, in order to destroy the diapause larvae, is not practicable in peasant agriculture as these residues are required for such purposes as palisading, building of contour banks and for grazing cattle and goats during the dry season.

Chemical control of the first generation of Busseola in young maize can be achieved by the application of a dust containing 2·5 per cent. DDT into the funnel of the plant. In 13 experiments on peasant-owned plots at Mbeya, in which a dust containing 2·5 per cent. DDT was applied four times, at weekly intervals, beginning about 18 days after germination, and at the rate of 10 lb. per acre, the average weight of maize cobs from 100 plants was increased from 12·46 to 22·82 kg., an increase of 83·1±19·0 per cent. Factors other than Busseola which affect the yield are discussed and it is suggested that control measures would probably be worthwhile only where a combination of a suitably fertile soil, early planting and adequately distributed rainfall obtains.

The quantitative study of Tyroglyphid populations involves their separation from the food material, for which existing methods are not entirely satisfactory A new method is described which depends on chemical treatment to remove the bulk of the flour. It is made quantitative and independent of moisture-content variation by application of the lycopodium-spore method of Wallis. A techniqul for determining the numbers of hatched and unhatched eggs per unit weight of flour is described. It provides information on the rate of increase or decreas in the number of eggs (natality rate per unit time) and hence on the ability the species to survive in a particular environment. The egg distribution in an undisturbed culture of Tyroglyphus farinae (Deg.) is found to be uneven. The technique is applied to the study of a population of Thyreophagus entomophagus (Lab.) under controlled conditions of temperature and humidity and the results discussed.

Estimations have been made of the weight of material accumulated by adults of Phormia (Protophormia) terraenovae R.-D. from deposits of DDT and other lipophilic particles on different types of substrate on which the flies were exposed.

Uptake of dry DDT particles is shown to be similar to that of dry lipophilic dye particles of similar size so that conclusions on the contamination of insects by dry lipophilic dyestuff may be extended to contamination by DDT dust.

The presence of oil greatly reduces the uptake of DDT particles from a fibrous substrate, the uptake decreasing logarithmically with increase in the amount of oil applied to the substrate. The viscosity of the oil has no more than a very small effect on the uptake of lipophilic particles from oily deposits and the solubility of the particles in the oil has no significant effect.

A reduction in the uptake of dry lipophilic particles from filter paper and from a hard, waxy substrate occurs when a wetting agent solution is applied immediately before the particles. When the particles are applied after the drying of the wetting agent on the substrate, a reduction in uptake also occurs on the non-absorbent, waxy surface, but is not so large as when the particles are applied before drying of the wetting agent. No reduction in uptake occurs on the absorbent substrate under these conditions.

On the basis of the experimental results the variations in the adhesion of lipophilic particles to substrates are discussed.

While differences in the affinity of particles for substrates of different types cause variations in adhesion, the physical structure of the surface may be more important in determining the amount picked up from it. The presence of oil or wetting agent on the surface increases adhesion, probably by increasing both the affinity of the lipophilic particles for the substrate and the area of contact between the particles and the substrate.

The quantities of lipophilic dyestuff picked up by specimens of five insect species, namely Vespula vulgaris (L.), Feronia madida (F.), Rhagonycha fulva (Scop.), Dysdercus fasciatus Sign, and Notostira erratica (L.), in their first few steps on a dust deposit of the dyestuff have been estimated. The uptakes per step by insects of different size and structure were shown to differ only slightly. While the uptake was found to be greater for the larger insects, the smaller insects picked up more particles per unit body weight at each step.

The ratios of the actual amounts (estimated by a colorimetric method) picked up by the different pairs of legs are similar for all the insects that have been studied and in all cases the hind legs were shown to pick up an amount at least equal to the sum picked up by the fore- and middle legs.

This paper presents the results of an investigation of certain processes taking place when flies of the species Phormia (Protophormia) terraenovae R.-D. are exposed to finely particulate deposits. Information is advanced concerning the precise sites of uptake of particles on the tarsi and pretarsi, the effects of “cleaning” reflex movements on the location and retention of particles and the influence of lipoid solubility (a fundamental property of contact insecticides) on the adhesion of dry particles to the lipophilic surfaces of insect and leaf cuticle.

A technique for the controlled exposure of active blowflies of the species P. terraenovae to particulate deposits is described.

The population of the Makonde plateau, Newala district, Tanganyika, underwent a severe epidemic of a virus disease in 1952–55, and during an investigation of the epidemic the opportunity was taken to define the biting activity cycle of Aëdes (Stegomyia) aegypti (L.), a mosquito that was extremely common in native huts on the plateau.

Continuous baited catches, lasting for 49 hours, were made by teams of four native catchers, inside a hut and on its verandah. The intensity of attack is expressed as the geometric-mean numbers of A. aegypti alighting on such a team per hour. The catches are recorded on a 24-hour system in which sunset is near to 1800 hr.

The responses of females of Aëdes aegypti (L.) to beef blood have been investigated. Paired comparisons were made of the numbers of mosquitos attracted to test and control materials.

Vapours emanating from the heparinised blood were found to be significantly more attractive than water vapour in an olfactometer. When the blood vapour was passed through an aqueous solution of Ba(OH)2 this attractiveness disappeared.

Exposure of heparinised blood on filter paper proved it to be two to four times as attractive as water or washed corpuscles suspended in saline. Plasma proved to be 4·7 times as attractive as water, indicating that the attraction of blood lies entirely in the plasma.

Treatment of blood with BaCl2 and a vacuum, thus rendering it incapable of evolving CO2, removed its attractiveness to mosquitos. Treatment with vacuum alone, which may be expected to remove the volatile substances, did not completely remove the attractiveness; nor did the treatment with BaCl2 alone, which removes the bicarbonates in the blood.

The attractiveness of blood was completely removed by desiccation and was not returned by remoistening. No attractive substance was isolated in an ether extract of blood.

When CO2-freed blood was treated with gaseous CO2 its attractiveness was significantly increased. When the treatment was such as to restore the normal CO2-producing capacity, the attractiveness was one-half that of whole blood. The attractiveness of whole blood was not increased by adding excess CO2.

It is concluded that the CO2 in beef blood is a definite element in its attractiveness, though probably not the most important one.

The present world distribution of six species of Cryptolestes found in stored produce is delineated from the evidence of specimens collected from a small number of countries and from specimens collected from produce examined in ships at British ports.

It is concluded that C. ferrugineus (Steph.) is worldwide, being found in temperate and tropical areas and in humid and dry zones. The range of C. minutus (Ol.) is limited by low temperature and low humidity but it is more abundant than C. ferrugineus in the wet tropics. C. pusilloides (Steel & Howe) is at present confined to the southern hemisphere but may be spreading slowly. It seems unlikely that this species entered stored products handled by the shipping trade much before 1944. C. turcicus (Grouv.) has been identified from Europe, North and tropical Africa, Turkey, North America, Uruguay and Argentina, and C. spartii (Curt.) only from Europe, North Africa and Turkey. C. ugandae Steel & Howe is rare, being found only from Uganda, Nigeria and Ghana.

Field and laboratory studies on the water-velocity preferences of the larvae of Simulium monticola Fried, and S. variegatum (Mg.) show that both species occur in velocities of 0·5–2·5 m./sec. but aggregate in the velocity range 1·0–2·0 m./sec.

The velocity range that the larvae of different species of Simulium can withstand indicates that current speed plays an important part in governing the distribution of blackfly larvae, and perhaps isolating different species along watercourses.

The laboratory rearing of all stages in the life-history of the Wheat Bulb Fly, Leptohylemyia coarctata, (Fall.), is described. The object of this work was to provide suitable material, in particular eggs and larvae, for insecticide experiments. Details of a container suitable for ovipositing adults are given. Several diets for the adults were compared and it was found that the greatest number of eggs was produced when the flies were fed on beef blood and a mixture of sweetened condensed milk and honey. Females were found to lay up to 244 eggs each; egg-laying commences 9–15 days after emergence and continues almost as long as the females live, but reaches a peak 3–5 weeks after emergence. Some females survived from late June until early November. In a small laboratory population, the average number of eggs laid by each female was 29·9 and mean length of life was 26·3 days for males and 41·6 days for females.

The eggs have a long obligatory diapause period. They have no effective water-proofing mechanism, but can recover from considerable temporary desiccation. A method of storage during diapause is described. The eggs normally hatch in late January or early February, but hatching can be delayed until much later in the year by storage below 0°C.

The effects of temperature on larval and pupal development are described. Larvae will grow at 20°C., but pupating larvae or young pupae are rendered sterile unless kept at temperatures of 15°C. or below. It is shown how a combination of delayed hatching and rapid larval and pupal development can produce fertile flies at unusual times of the year in the laboratory. It is suggested that this may make it possible to produce young larvae in the autumn for insecticide experiments.

In conclusion, some parasites and predators encountered in the field and the laboratory are mentioned.

The loss of crop following insect attack on cotton was studied in 1950–54 at the Cotton Research Station, Namulonge, which is situated in the elephant-grass zone of Uganda, in which the principal pest of cotton was originally considered to be Lygus vosseleri Popp. In 1953, the main invasion of cotton by Lygus occurred during September and October, originating from cultivated crops of black gram (Phaseolus mungo) and sorghum and from the wild perennial legume, Pseudarthria sp., which earlier workers had considered important. Trials in this zone have demonstrated, over a number of years, that early June is the optimum date for sowing cotton, although such sowings receive the heaviest attack by L. vosseleri. Experiments to determine the maximum loss of yield due to Lygus should therefore be made with cotton sown at this time. Bollworms are also important pests, notably Heliothis armigera (Hb.) and Earias spp., their attacks following those of Lygus and being heaviest in October–November.

The study of crop loss following insect attack on cotton in Uganda, described in Part I of this work, was continued for two further seasons (1954–56), using the same experimental procedure, except that half the DDT in the spray formulation was replaced by toxaphene, only three periods of protective treatment were used for each date of sowing, and the late-season ones were continued until boll-opening was complete, so as to give maximum control of boll worm.

Methods are described of estimating an index of leaf damage caused by Lygus spp., based on comparisons with standard grades. At Serere (Eastern Province), the index of leaf damage was closely related to Lygus populations in the crop, and at Namulonge (Buganda) it was markedly reduced on plots protected from Lygus attack, but in the 1954–56 experiments no relationship could be found between the mean seasonal index and the loss of fruiting bodies attributable to Lygus attack, or the final yield.

Analysis of the cause and incidence of shedding of developing fruit bodies (flower buds or young bolls, the latter predominating among those recovered) showed that the total amount of shedding was roughly the same whatever the treatment, that due to insect damage tending to be balanced by that due to other causes, which were probably physiological in nature. There was thus some over-riding factor, other than pest attack, that governed the crop finally held by the plant.

Damage caused by insects to older bolls, which do not shed, constitutes a real loss that is reflected in the final harvest. The proportion of clean seed cotton was higher in the treated plots, particularly in the early-sown cotton, in which analysis of the harvested locules showed that the damage was predominantly due to bollworm, but there were no significant differences in yield between the various treatments in either date of sowing or in either season.

It is concluded that under such conditions as those experienced at the Research Station in the three seasons of the experiments, insect damage to cotton is small in relation to other environmental factors governing crop production.