Wood-and-fibre-board huts fitted with one-way entry and exit traps were sprayed with DDT and BHC wettable powders at dosages 3·3 and 1·5 times the dosages reported by Muirhead Thomson in Tanganyika. Mud huts similar to his design were sprayed at 1·4 times his dosage of DDT and 1·1 times his dosage of BHC. Catches of Anopheline mosquitos continued for 16 weeks.

In the first-named huts kills of A. gambiae were 97 per cent. for both treatments. Kills of A. funestus were 100 per cent. and 97 per cent. for DDT and BHC respectively.

In the mud huts kills of A. gambiae were 62 per cent. and 98 per cent. for DDT and BHC respectively. For A. funestus kills were 78 per cent. and 93 per cent. in the same order.

The advantages claimed for the wood-and-fibre-board huts include :—

(a) freedom from ants, which may remove dead mosquitos.

(b) ease of catching live adult mosquitos.

(c) rigid construction of exit trap.

(d) lack of adventitious exits other than exit trap.

1. A warm object is more attractive to the female yellow-fever mosquito than a cooler object, such that the numbers touching a ball warmed to 100° or 110°F. were twice as great as those touching a ball 20°F. cooler. This attractiveness was reversed when the temperature of the warmer ball reached 120°F.

2. This response is eliminated by the insertion of an air-tight window of thallium bromoiodide, despite the fact that it allows almost all of the radiation to filter through. It is therefore concluded that heat convection, which is eliminated by the air-tight window, is the factor which makes a warm object attractive to the mosquito.

3. The several faces of a warm cube, differing in their radiant emissivity but identical as to surface temperature, exhibited attractancies to Aëdes which were not significantly different one from another (except in the case of black enamel, which will be explained in a later paper). These results may be interpreted to indicate that radiation plays no part in the attraction of the mosquitos to the warm face.

4. It is concluded that the failure of Parker (1948) to obtain a positive response of Aëdes mosquitos to a warm dry object was due to the fact that his apparatus allowed only the radiant heat but not the convective heat to reach the insects.

5 Since the response of Aëdes aegypti to heat is eliminated by a window transparent to infra-red radiation, the theory of olfaction proposed by Beck and Miles (1947) evidently has no application to this species of mosquito.

A number of leaf-eating insects have been studied with the object of selecting species suitable for rearing in the laboratory.

Four species have proved satisfactory for large-scale rearing throughout the year. These are Plutella maculipennis, Plusia gamma, Diataraxia oleracea and Phaedon cochleariae. A number of species possess certain disadvantages : (1) apparently obligate diapause (Mamestra brassicae, Sphinx ligustri). (2) The same as (1) with the additional disadvantage of lack of food-plant in winter (Lymantria dispar). (3) Uncertain copulation and oviposition (Pieris brassicae). When available, these species may be reared satisfactorily in large numbers.

Mesographe forficalis, Xanthorhoë fluctuata, Lema melanopa and Crioceris asparagi have been reared in relatively small numbers. They possess certain disadvantages but should be satisfactory for large scale rearing at any rate during some part of the year.

At present laboratory rearing of Phlogophora meticulosa, Agrotis segetum, Triphaena pronuba, Aclypea opaca and Athalia colibri is considered impracticable. A number of leaf-eating species, particularly Hymenoptera, which appear to possess suitable qualities for laboratory rearing, have not yet been obtained and studied in the laboratory.

Particular attention has been paid to factors influencing development of leaffeeding insects with the object of determining the conditions under which healthy stocks of uniform resistance can be maintained. The results of experiments on the effect of food-plant, temperature, humidity and illumination are described. Studies on the factors influencing diapause in the various species are mentioned. Descriptions, are given of the rearing techniques that have been developed.

The results of a series of catches of biting insects at human bait on platforms in the canopy of the forest at Mongiro, Bwamba, Uganda, are reported. The catches were designed to investigate differences in the numbers taken, due to platforms, season and weather, but mainly to study the short-term changes in the activity of biting insects near the time of sunset. Each catch lasted from 3 hours before to 3 hours after sunset, and was divided into 10-minute periods from 30 minutes before to 1 hour after sunset. The main analysis is of catches made on five platforms on each of 100 evenings which covered the end of the dry and beginning and end of the wet season, 1948, and the end of the dry season, 1949.

A list of the species of biting insects encountered is given. Eight species (Anopheles (M.) gambiae, Taeniorhynchus (C.) fuscopennatus, Taeniorhynchus (M.) africanus and uniformis, Aëdes (M.) nigerrimus, Aëdes (S.) apicoargenteus and africanus, and Chrysops centurionis) were found to be abundant and data relating to them has been subject to more or less detailed analysis. The remaining 27 species or groups of species are discussed shortly.

Six of the eight abundant species show statistically significant differences between the catches on the five separate platforms. Arrangement of the five platforms in order of the total numbers taken, for each of these six species, gives results which are practically identical in three species (Anopheles (M.) gambiae, T. (M.) africanus and uniformis) and in two others (Aëdes (S.) apicoargenteus and africanus) ; however, the arrangements in these two species groups differ widely from each other. The arrangement for C. centurionis differs again.

These ordinal arrangements are compared with the obvious characteristics of the platforms and no consistent relationship is found. It is pointed out that the first three species are derived, largely or even entirely, from breeding places outside the forest, in or associated with the adjacent extensive areas of open swamp, while the last three are likely to be completely confined to the forest for breeding ; both the latter two mosquito species breed in tree holes, and dense shade is characteristic of the breeding places of species of Chrysops closely allied to C. centurionis. It is considered most probable that these platform differences in the numbers yielded are due to local differences in the densities of the adult populations which in turn are related to the abundance and proximity of breeding places.

Seasonal variations in the numbers of the abundant species are also studied and are considered to be almost certainly related to the amount and distribution of rainfall affecting the profusion of breeding places at least as far as the mosquito species are concerned. The tree-hole breeding species, however, suffered a reduction in numbers before the end of the wet season, an effect also noted by Mattingly in West Africa.

The amount of the reduction of the numbers of A. (S.) africanus in the dry season is of importance in assessing the possibilities of the survival of yellow fever virus in that mosquito as a host. The estimated reduction in the dry season of 1949 was to less than one twenty-third of the population existing at the end of the previous wet season (November, 1948).

C. centurionis occurred during all the catch series but was extraordinarily abundant in late March and early April, 1948. It is considered probable that this time of increase represents a yearly period into which hatching from the pupa is compressed. This conclusion is supported by the finding that during this period the biting population was composed of two groups one of which tended to bite before, and the other after, sunset.

Evidence is produced to show that the biting activity of Anopheles (M.) gambiae was depressed by wind, that of Aëdes (S.) africanus by wind and rain but that that of C. centurionis was little affected by either of these factors.

Studies of the biting-activity changes in the nine ten-minute periods extending from 30 minutes before to one hour after, sunset are presented. Each of the species which were known to be mainly active by night and which were sufficiently abundant for study, show activity increasing about sunset, rising to a maximum shortly after, and thereafter declining again to moderate levels. This initial peak is very clearly marked in the cases of Anopheles (M.) gambiae, T. (C.) fuscopennatus, T. (M.) africanus, Aedes (S.) africanus and C. centurionis, less so, but still distinct, in T. (M.) uniformis and A. (M.) nigerrimus. Its position differs in different species being earliest in the case of C. centurionis (18.10–18.20 hours) and latest in that of T. (M.) africanus (18.40–18.50 hours). However, in any given species the peak period of biting activity was subject to only slight movement in time in relation to sunset; in analyses of the data by different platforms, by season, and by types of weather it was the exception to find the time of peak activity to be shifted more than one ten-minute period away from the position which it occupied over the whole series. Further, in studies of the curves of biting activity on individual nights, three consecutive ten-minute periods centred on the period which showed the peak over the whole series, included the peak in 21 of the 23 A. (M.) gambiae cases available, in all five of the T. (M.) africanus, in eight of the 11 A. (S.) africanus and in 15 of the 20 C. centurionis, cases. The regularity of the occurrence of the peak period of activity in any given species is considered to indicate that it is largely controlled by changes in light intensity about the time of sunset.

These findings are discussed in relation to the general problem of the mechanism controlling the rhythm of activity in insects biting by night and a hypothesis, which is based on the conception that different population groups may have different biting habits, is advanced to account for the different types of cycle commonly encountered.

A block of a mile and a half of riverine forest on the Kuja River was separated from other riverine bush by clearings of 2 and 3 miles. Clearing costs are recorded.

An attempt was made to exterminate G. palpalis in the block by hand-catching. This failed, because the clearings did not exclude immigrant G. palpalis.

More G. palpalis could be caught from boats on the river than from paths beside the river.

Two attempts were made to prevent G. palpalis from crossing the clearings :—

(a) by stretching hessian screens treated with DDT right across the river.

(b) by placing large numbers of DDT treated “ trap-screens ” in the clearings. Neither had a decisive effect.

No evidence was found to indicate that hand-catching would not work in a completely isolated block, but it appears to be quite impracticable to isolate a block of G. palpalis bush on a permanent river.

Methods of preparing solutions of graded density for the accurate control of atmospheric relative humidity are described, and some pitfalls in their use and in the use of saturated salt solutions are indicated.

For graded solutions of potassium hydroxide and of sulphuric acid, data from the International Critical Tables or more recent sources are used as the basis of tables giving the concentrations (wt.%) and densities corresponding to relative humidities in steps of 5 per cent. R.H. Sources of similar data for calcium chloride, sodium hydroxide, sodium chloride, and glycerol solutions are given.

As an addition to the compilation of the available data on humidities in contact with various saturated salt solutions by O'Brien (1948), some more recent figures are quoted.

1. Flies of 9–10 and 16–19 days old were more susceptible to DDT than 4–5-day-old flies. Female flies were less susceptible than males, and large flies less susceptible than small flies to residual deposits of DDT. It is clearly important that flies used in comparative toxicity tests should be standardised for age, sex and size.

2. The strain Torre in Pietra was 3·6 times more resistant to dry deposits of DDT on Essex board than the Roma strain, but this resistance declined after culturing the strain in the laboratory for 22 generations.

3. The DDT-resistant strain (Torre in Pietra) was as susceptible as the non-resistant (Roma) strain to deposits of γ BHC on Essex board but was slightly resistant to pyrethrins.

4. The DDT resistance of the Torre in Pietra strain was increased readily by selection, yet a similar selection with DDT on the Roma strain was unsuccessful. This failure in the Roma strain may be due to the purity of the stock.

Living specimens of Megarhinus brevipalpis were transported from southern Natal to Johannesburg to establish an insectary-bred colony. The natural habitat of these predatory mosquitos consisted of small isolated patches of sub-tropical forest, in which the rainfall is 40–50 ins. (102–127 cm.) with a mean winter temperature of 64°F. (17·7°C.) and an annual range of 27°–33°F. (15°–18°C). The breeding places were leaf axils of Strelitzia nicolai (a plant resembling a wild banana), small rot holes in trees and larger ones in Strelitzia stumps. The larvae were collected from leaf axils with an apparatus consisting of a rubber bulb to which were attached lengths of glass and rubber tubing.

The insectary was a room 9 ft.×8 ft. 6 ins. and 9 ft. high which was kept at tropical heat and humidity. Mating of the adults was observed, copulation being effected while at rest or in flight. Oviposition was usually accomplished in flight but also while at rest on the surface of the water. In the summer time two females, which were tested, laid about 85 eggs each during the month following emergence from the pupa, six or seven days elapsing after emergence before the first oviposition. In the middle of the winter, oviposition (with later generations) became very irregular in spite of the temperature and humidity remaining constant. The adults, which were comparable to those of the natural habitat, were fed on sugar solution, honey and fruit juice. One bred out as a gynandromorph.

When given an abundant supply of larvae of laboratory bred Aëdes aegypti, the life-cycle of M. brevipalpis was normally : egg (incubation), less than two days ; larva, 11–20 days (average 14·5 days) ; pupa, five days. This does not include a small number of exceptional cases in which the life as a fully grown larva was abnormally prolonged (in one case nearly four months) for reasons which are not absolutely clear. The larvae killed from 100 to 200 or more Aëdes larvae during the normal larval life, but many of these were not eaten when the brevipalpis were in the late fourth instar. By a special technique they were also induced to eat dead tissues including minced pork brawn, minced maggots and minced flies. Except for the latter these were not satisfactory foods although there was slow development.

Fourth-instar larvae were kept out of water for three to four weeks (without food), in a damp atmosphere, and afterwards when fed most of them developed normally, but pupation was sometimes suspended for a considerable time. They have been sent by post (out of water) in tubes with damp cotton wool and filter paper.

The egg differed from that of other Megarhinus species in having a crown of projections at one end with a cup-like structure in the centre. The exochorion had roughly hexagonal cells but without numerous tubercles as in other species.

First-instar larvae remained in the egg-shell after hatching when the eggs-were out of water but on a damp surface and in a saturated atmosphere. They survived like this for up to six days or about the same time as the larvae survived in tap water if there was no food. When liberated in water the head of the first-instar larva was comparatively small with the mouth parts folded in. Within two hours of liberation in water the head enlarged considerably and the mouth parts came into position ; the larva was then ready to catch its Culicine prey. When in water containing dead leaves, these larvae survived from a few days to over four weeks and some grew to the third instar without any Culicine food.

Cannibalism was investigated. Fourth-instar larvae did not attack each other readily ; they devoured smaller larvae of their own species and small to medium size larvae resorted to cannibalism, particularly in the absence of Culicine prey. There was evidence that fourth-instar Aëdes aegypti occasionally ate first-instar Megarhinus.

The discussion traces attempts which have been made in certain Pacific islands, notably Hawaii and Fiji, to use Megarhines for biological control of disease-carrying mosquitos. M. brevipalpis has a shorter life-cycle than the species introduced into these islands and the conclusion reached is that laboratory breeding, to enable large numbers to be released in certain areas, would be a suitable adjunct to a programme of general control, in this part of the world. Airmail consignments of larvae are being sent to Hawaii with the object of starting a laboratory colony there.

Over 3,000 examples of G. palpalis were age-grouped and dissected for T. vivax over a period of a year. The incidence of infection throughout was highest in Age-group 3 and lowest in Age-group 1. There was a marked seasonal fluctuation, the incidence being highest during the months of heavy rainfall, i.e., August to October, and lowest in the driest.

It has been suggested that the rise in the rate of infection during the rains is merely due to the increasing average age of the fly population, but by the method of age grouping it is shown that this is not the case.

The degree of infection in G. palpalis in Sierra Leone is comparable with that of G. tachinoides in Nigeria.

It is not thought that the seasonal variation in the incidence of infection can be ascribed to changes in the type and abundance of food. The true cause of the fluctuating incidence of infection like the source of infection is unknown.

The failure to recognize seasonal differences in the rate of infection has led to some important practical misconceptions.

The weights of insecticide picked up by mosquitos and tsetse flies in short contact periods on deposits from aqueous suspensions of DDT crystals of different sizes have been determined chemically and correlated with corresponding mortality rates.

The rate of action of DDT particles picked up by mosquitos increases as the size decreases.

Taking into account the weights of insecticide picked up, the median lethal dosage increases as the particle size increases. A further factor, that of retention on the body of the insects, has to be considered, however, for there is evidence that adherence of particles in the 0–10 and 10–20 micron ranges is greater than that of larger particles. Availability, and therefore effectiveness, of deposits from aqueous suspensions of insecticide particles is influenced considerably by the type of material to which they are applied.

A method for the routine quantitative estimation of mealybug populations on cacao trees is described and the results obtained from 24 successive monthly samples are tabulated.

The quantitative results obtained are of such variability that only gross differences in population density are shown as statistically significant.

From a series of analyses based on the survey data the following conclusions have been reached with regard to P. njalensis, the most important mealybug vector of swollen shoot virus :—

(a) The species is almost invariably attended in the field by Crematogasterine ants, the most important ant species building protective carton tents over the mealybug colonies.

(b) Crematogaster density is closely correlated with mealybug density and the ants construct their carton tents in relation to the numbers of mealybugs to be protected. Some of the arboreal Crematogasterine groups investigated are shown to be consistently associated with higher mealybug populations, and to have a more advanced tent building habit, than others.

(c) Tree to tree variation in mealybug density is shown to be largely dependent on the identity of the dominant ant group.

(d) No direct correlation is apparent between mealybug density and incidence of swollen shoot though populations in areas devastated by virus are usually smaller than in areas in which virus is actively spreading.

(e) The factors of importance in the natural control of P. njalensis populations have been discussed in some detail. It is clear that the species is maintained by ant protection at a density level considerably above that prevailing amongst closely related pseudococcid species which are not ant attended to the same extent. At this level of “ protected steady density ” balance is maintained partly by losses during crop harvesting, and partly by swollen shoot rendering the host plants unsuitable as feeding stations. It is concluded, however, that natural enemies must, in the last analysis, be responsible for maintaining balance.

(f) Since the protected density of P. njalensis is a direct result of ant associations it follows that factors tending to control the attendant ant species will have a delayed effect on mealybug density. In this respect it is believed that Oecophylla, which is strongly negatively correlated with Crematogaster spp., is of importance. Any attempts that may in the future be made to control the mealybugs by killing the ants will have to be designed as specific against the Crematogasterini and non-lethal to Oecophylla and the other large predatory ants common on cacao.

Conwentzia psociformis (Curt.) and Conwentzia pineticola End. differ as adults in only two characters, and since there are variable and intermediate forms occur, reliable determination is sometimes impossible. In the past, these two species have often been regarded as two forms of the same species.

The immature stages of C. pineticola are, however, very different from those of C. psociformis and show that the two are distinct species and can readily be separated on larval characters. The larval stages and egg of C. pineticola are described for the first time, and an account of its life history given.

The feeding habits of these two species are described, and mention is made of two other Coniopterygid species that occur in orchards.

An investigation on the taxonomy of A. stephensi and A. superpictus larvae in Iraq is reported upon.

As a result it is suggested that 90 per cent. of the larvae of these species can be separated on the relative length of the filaments of the abdominal palmate hairs.

The question of the attractancy of water-vapour, carbon dioxide and candidate attractant compounds was investigated by the use of an olfactometer emitting freely into a 360-cu. ft. cage filled with female adult Aëdes aegypti. Moist air (85 per cent. R.H.) was found to attract 3–5 times as many mosquitos as dry air (15 per cent. R.H.). A moistened billiard-ball was found to attract about seven times as many mosquitos as a dry one, the result being similar whether black or white balls were used. The attractancy was reduced and even reversed at very high room humidities.

The addition of 10 per cent. CO2 to a dry air stream approximately doubled its ability to bring mosquitos within 0·5 inches of the olfactometer port. Its addition to a moist air-stream did not significantly increase its attractancy. Pure CO2 vapour proved to be no more attractive than dry air, and several times less attractive than moist air.

A number of chemical compounds found in body exudations, and some housefly attractants, were tested in the olfactometer for their effect on Aëdes adults. None were consistently attractive, and many were significantly repellent. However, sweat itself proved significantly attractive at a low vapour concentration, while significantly repellent at a high concentration.

Owing to the courtesy of Miss H. Dombrowski (E. V. Dombrovskaya) of the Zoological Institute of the Academy of Science, Leningrad, U.S.S.R., it is now possible to describe a sixth species of gall midge belonging to the genus Trisopsis Kieffer. Males, females and larvae of this species were received from Russia in April, 1937. The larvae had been found during the previous winter preying on the Flour Mite, Tyroglyphus farinae (Deg.) during laboratory observations on the predaceous mite, Cheyletus eruditus (Schr.), which had been kept with the flour mites in paraffin wax cells at the Institute for Plant Protection, Leningrad (Sorokin, 1938).

Eurygaster integriceps is a serious pest of wheat in Syria against which various methods of control have been tried.

Laboratory tests with a 0·8 per cent. parathion dust applied at the rate of 0·1 mg./sq. cm. are described.

Three sets of field experiments were carried out with 0·75 per cent. and 0·8 per cent. parathion dust. Applied at a rate of 75 kg. /ha. these were very efficient but not economical. Applications of 0·8 per cent. dust with a knapsack duster at between 15 and 17·5 kg. /ha. gave about a 20 per cent. reduction in the infestation. A larger experiment in which 0·75 per cent. dust at a rate of 48 kg. /ha. was applied to a hectare of wheat, by means of a 6 h.p. duster mounted on a Jeep fitted with a drag sheet, gave a 70 per cent. control.

A number of liquid mosquito repellents were assessed for vapour repellency power in an olfactometer mounted in a very large cage filled with females of Aëdes aegypti. They were also tested for their knockdown power in fumigation bottles. Their vapour pressures were determined by the Ramsay-Young method.

All the liquids showed vapour repellency, and in 39 out of the 42 tested this effect was highly significant. The highest vapour repellency ratings were shown by compounds already known to be the most effective repellents.

Although the more volatile compounds such as citronellal tend to show the highest repellency ratings, nevertheless compounds of low vapour pressure such as indalone, DMP and isobornyl morpholinoacetate may also show high vapour repellency. It is concluded that vapour repellency, although in the first instance dependent upon volatility, can vary independently of vapour pressure, so that compounds may be found which afford not only a long protection period due to their nonvolatility, but also a high vapour repellency due to the potency of the comparatively few molecules that are volatilised.

The vapours of most of the repellents were found to induce knockdown of mosquitos, but there was no correlation between the speed of this process and the vapour repellency of the compounds.

Whilst the status of Capnodis as a pest in Israel has been considerably reduced by the control measures that have been adopted, it was considered desirable to test the new synthetic contact insecticides, in particular DDT and BHC.

Techniques for the determination of the effect of insecticides on neonate larvae of Capnodis are described and the results discussed.

It is shown that BHC and DDT are ineffective as fumigants against the boring larvae.

Used as a contact insecticide against the larvae, a high degree of mortality is obtained with 5 per cent. BHC over a period of 50 days. DDT is far less effective. BHC, dieldrin and methoxychlor remained effective for over a year if left undisturbed in the ground.

Against adult Capnodis, BHC gave results comparable to those obtained with cryolite and retained its activity for at least twelve days after application.

Records are given of the localities (with altitudes) and habitats of 11 species of Diatraea Guild., all reared from known food-plants in northern Venezuela. The larvae of several of these moths are important pests of sugar-cane and maize. Among the eight previously described forms, D. busckella rosa Heinr. is here raised to specific rank, and D. busckella Dyar & Heinr. is treated as a polytypic species, in which typical subsp. busckella has three well-defined forms, two of them, viz., f. falconensis nov. and f. setariaeoides nov., being described, while others are indicated ; the subsp. setariae nov. is also described. The diagnostic characters of D. rosa and D. busckella are discussed and their genitalia are illustrated. The various subspecies and forms of D. busckella are discussed in relation to each other and to D. rosa, these two species being considered as together comprising a supraspecies. Hybrids have been reared in the laboratory among certain of the components.

Three new species, D. andina, D. pittieri and D. silvicola, are described and illustrated, all from northern Venezuela.