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.

Following increased yields of cotton in the Sudan Gezira by DDT spraying, during the 1952/53 season 22,300 and 6,700 feddans of X1730A cotton were sprayed once and twice, respectively, in the Gash Delta of eastern Sudan, where nearly 60,000 feddans were grown under controlled flush irrigation. Each spray consisted of 1 lb. technical DDT per feddan and was applied by aircraft in 2 gals, of spray per feddan.

Systematic observations were made on the incidence of cotton pests in 30 observation stations scattered throughout the Gash Delta, such stations being selected more or less at random from cotton sprayed once and twice and unsprayed, ten amongst each of these three treatments.

A single DDT spray applied 50–70 days after sowing gave entirely satisfactory control of the Jassid, Empoasca lybica (de Berg.), and the thrips, Hercothrips fumipennis Bagn. & Cam. and H. sudanensis Bagn. & Cam., throughout the growth of the crop. There was little lasting control of the flea-beetle, Podagrica puncticollis Weise. The second spray, applied 70–90 days later, had little effect on any of these pests which were then present in low numbers.

The incidence of bollworms was observed between December and March, covering the important fruiting period. During these months, over 70 per cent, more larvae of Diparopsis watersi (Roths.) and 40 per cent, more larvae of Earias insulana (Boisd.) were observed on sprayed than on unsprayed cotton. Significantly more larvae of D. watersi were recorded on cotton sprayed twice than sprayed once; conversely, significantly fewer larvae of E. insulana were recorded on twice- than on once-sprayed cotton. Moreover the estimated total number of fruits damaged by or shed in association with bollworm attack between January and March was over 30 per cent, greater, and nearly 30 per cent, more bollworm damage to nearly mature bolls was recorded between December and March, on sprayed than on unsprayed cotton.

Yields of seed cotton were significantly less from twice-sprayed than from the other treatments. Once-sprayed cotton yielded less but not significantly so, than unsprayed cotton. Yield was negatively correlated with the number of sprays but the correlation coefficient just failed to reach significance.

Further analysis of the data from stations where yield differences were most marked indicated that D. watersi lowered the yield potential of the crop as expressed by fruit production and retention, as a result of continuous shedding of the damaged fruit. Attack, however, was concentrated on cotton with the best yield potential. The effect of spraying was to increase the numbers of D. watersi and thus to give rise to excessive shedding, increased production of fruit primordia, an increased number of damaged bolls, and finally reduced yield.

It is concluded that any benefits which the crop enjoyed, as a result of elimination of leaf-feeding insects by DDT spray, were completely lost through increased bollworm attack, which moreover reduced the yield below that of unsprayed cotton.

Experiments were carried out in Kenya to find whether insecticidal smokes from “Swingfog” machines would remove tsetse flies, in this instance almost all Glossina longipennis Corti, carried by trains. Fogging was done by two operators, each with a machine, starting from opposite ends of trains halted at a station twenty miles inside a fly-infested area and assessment was by routine catches at a station twenty miles outside it. It was found that fogging reduced the number of flies found on trains by 60 to 70 per cent. As the trains had to pass through a light fly infestation after fogging, the real kill is likely to have been higher. The cost of such partial reduction is believed to be considerably higher than would be that of therapeutic control of trypanosomiasis in the affected region.

The validity of the MLD (Minimum Lethal Dose) as a practical concept for measuring toxicity, despite its obvious lack of precision, is discussed. Its use has justification.

The 48-hour MLD of acetone-water suspensions and oil films of DDT and dieldrin for early fourth-stage larvae of Anopheles gambiae Giles have been measured.

The results show A. gambiae to be relatively resistant to DDT.

In acetone-water suspension dieldrin appears to be about six times as toxic as DDT. In oil solutions this difference is more marked, the ratio being in the neighbourhood of 20:1.

There is a clear diurnal rhythm in the emergence of the Beet Fly, Pegomyia betae (Curt.), from the soil, peak emergence occurring daily between 06. and 07. hr., when the soil temperature at a depth of 2 in. and the air temperature are at about their minimum, or are just beginning to rise, and humidity is at its maximum. In two observations on the emergence of the first generation, 82·2 and 85·6 per cent, of the daily emergence occurred before 08. hr., whilst in two observations on the emergence of the adults of the non-diapausing part of the second generation, 79·1 and 76·3 per cent, of the daily emergence occurred before 08. hr.

During the period of emergence of flies from a batch of puparia, the rhythm was most marked at the time of peak emergence, and was less marked at the beginning and end of the emergence period.

The rhythm of emergence was much less pronounced, and the peak occurred later in the day, when puparia were kept at constant temperature from two days before the first flies emerged.

Experiments led to the conclusion that the fly reaches the soil surface within one hour of leaving the puparium at a depth of 2 in. in soil, and that the processes of wing expansion and cuticle tanning are probably controlled by a nervous mechanism.

The possible initiators of the diurnal rhythm are discussed. It is concluded that the results obtained could be explained by the suggestion that the rhythm is induced by the effect of temperature or light fluctuation at some time before the late pupal stage, or that it might be inborn in the species, and that it is further regulated by temperature variation at the time of emergence. Further work will be necessary before the suggested explanation can be shown to be valid or otherwise.

Six new species of stem-boring AGROTIDAE are described from Africa. The new genera, Carelis, Poecopa and Manga are erected to accommodate three of these species; the other three species are described in the genus Busseola.

A key to nine genera is given, and also one to separate five extremely similar species, Poeonoma serrata (Hampson), B. fusca (Fuller), B. quadrata, sp. n., B. phaia, sp. n., and B. segeta, sp. n.

The Garden Chafer, Phyllopertha horticola (L.) is the chief British representative of a type of world-wide pest. The following summary concerns the adult and its reproduction under laboratory conditions.

Adult feeding affects neither the number of eggs produced by the ovaries (productivity) nor the number laid (fecundity). Productivity depends entirely on the amount of fat-body stored away by the third-instar larva before hibernation.

The adult stage lasts about three weeks (3 to 35 days) in both males and females. Starvation shortens the life by a day or two. Females which lay no eggs (barren) are usually short-lived, i.e., they die before oviposition is due to start, and comprise 14 per cent, of the population.

The fat-body is exhausted and the eggs are all matured by about the end of the first half of the adult stage. Oviposition and feeding apparently do not begin until the manufacture of eggs is completed. Individually, oviposition lasts five or six days (1 to 23 days), the more eggs laid, the longer does the oviposition period usually last. The female dies two or three days (0 to 10 days) later. Starved females have a rather shorter oviposition period. Unmated females may lay a few non-viable eggs just before they die.

The average female lays 13 eggs (0 to 46 eggs). More than 70 per cent, of females lay all or nearly all of the eggs manufactured from the fat-body. Apart from barren females, there is no significant difference in length of life between females that lay all of their eggs and those that lay only some.

Females are larger than males, being on the average 1·2 to 1·3 times heavier. Productivity and fecundity are positively correlated with female body weight (which is, of course, proportional to weight of fat-body). Thus regressions may be calculated such as:

F = 0·113W − 6·60

where F = number of eggs laid and W = weight of pupa in milligrammes. Female pupal weights vary from 80 to 310 mg. Body weight is not associated with the occurrence of barrenness. If all individuals always fulfilled themselves, a population would increase eight-fold annually.

There is not the slightest evidence of a fixed oviposition pattern. Four eggs may take nine days to lay while 27 may take only 24 hours. The same number of eggs may take a longer or shorter time to lay in larger or smaller driblets, the numbers of eggs and the intervals between them being apparently quite random.

A female burrows down at a different point each time she has eggs to lay. At each burrowing she deposits her eggs within a radius of one inch from the vertical line passing through her entrance hole on the soil surface. Eggs are laid singly in small cavities about quarter of an inch apart (3/16 to 1 in.).

Eggs hatch out in 4 to 6 weeks at temperatures of 15 to 17°C. Eggs failing to hatch vary between 4 and 28 per cent, of the total laid. They comprise both unfertilised eggs and, to a lesser extent, eggs which have been fertilised but die from other causes. The proportion of unfertilised eggs increases with the age of the female at oviposition.

A DDT sol, prepared in glass dishes in the manner described, has a reproducible zeta potential of −25 mv. Electrophoretic measurements showed that DDT conditioned in cardboard cartons has a zeta potential of zero.

The zeta potential of the suspension behaves in a predictable and measurable manner in the presence of mono-, di-, tri-, and tetravalent ions.

Spectrographic analyses showed that there are two identifiable sources of ions: (1) the DDT, and (2) the cardboard container. A third source of ions was indicated as being, singly or together, the distilled water and/or the DDT solvent.

Spectrographic data showed that ions were associated with the conditioned DDT.

In glass containers, a poor source of ions, kill of larvae of Anopheles quadrimaculatus Say was higher than in paper containers, a good source of ions. These data are correlated with the zeta potential and flocculation behaviour of DDT suspensions.

The Garden Chafer, Phyllopertha horticola (L.), has three larval instars, the third ending in hibernation, which gives way to a prepupal stage. The development cycle occupies 12 months and only one generation is present in the soil at any time. From a study extending over five years, 1948 to 1952, the cycle from egg to adult in the English Lake District may be outlined as follows:—

At the earliest, oviposition starts in the latter half of May. In soil 3½ in. deep, eggs are laid at an average depth of 1½ in. (range ¾ to 2½ in.). Other authors report that where the soil is sufficiently deep, the maximum egg depth may extend to four inches in the Lake District and even eight inches elsewhere. The eggs are spaced about a quarter of an inch apart (max. 1 in.), each in a tiny earthen cavity, all more or less directly below the point where the female enters the soil. Incubation of the individual egg averages five weeks.

The first instar occupies individually about three weeks on the average; the second instar about four weeks; and the third instar, up to the beginning of hibernation, eight to ten weeks. On the average also, first-instar larvae begin to appear in a population about the first week of July, second instars about the fourth week of July, third instars about the fourth week of August, and the earliest hibernators about the third week of October. Except for a few stragglers occasionally in early December, the entire population is generally hibernating by the end of November. The hibernation is a true diapause.

A detailed description of the method of feeding is given. The larva consumes plant roots which it obtains by tunnelling through the soil. Since its natural habitat is pasture land, grass roots are the main food. It probably also eats invertebrate carrion occurring by chance in its path. The first and second larval instars are given over to growth, the third and final is occupied mainly in storing up fat-body. This store has to suffice for maintenance during the remainder of development and also for the entire egg or sperm production.

On hatching, the larvae feed at about 1½ inches (¾ to 2½ in.) depth in the soil, i.e., at egg-level. As they grow, however, they ascend until latterly, as third instars, they are feeding about ½ inch (¼ to 1 in.) from the surface. This progressive rise is probably dictated by the increasing need for a more copious food supply. With the possible exception of a very prolonged drought, weather has no effect on the level at which larvae feed.

Larvae hibernate at 2 inches (0·8 to 3·8 in.) below the surface of soil 4 inches (2·3 to 6·0 in.) deep, i.e., well above the “ pan ” (gravel bed, or rock). In other localities other authors have recorded hibernation at the same as well as greater depth in deeper soils. The existing data are insufficient to show what governs the choice of depth in the soil.

On the average, prepupation begins in a population about the end of March and, individually, lasts between three and four weeks; pupation begins in the third week of April and lasts about four weeks. The pupa lies inside the last larval exuvium in the hibernation cell. Sex can easily be discerned in the pupa. There are always rather more males than females in a field population, considerably more in some years. Pupal sex ratios ranged from 1·13 to 1·89. On the average, male pupation precedes female by one day or a little more.

Behaviour after the splitting of the pupal skin is the same in male and female. At first, for about four days (1 to 7), the adult remains motionless in the hibernation cell. Then, alternately burrowing and resting, it ascends to the base of the sward in about two days (¼ to 4). At the sward base it now halts for about two days (½ to 6) before emerging into activity upon the sward surface for the first time; this halt of two days may be prolonged by one or more days if weather is unsuitable when a beetle is ready to become active. In toto, given no weather hindrance, the individual transit from cell to sward surface usually occupies rather more than one week (7·7 days, range 6·0 to 8·5). On the average, the first active beetles are seen in the last week of May but may be as early as the third week of May or as late as the second week of June, according to the weather. The male precedence over females in development is maintained from pupation onwards, hence the first males are always active upon the sward at least one day before the first females.

It takes some considerable time for all the individuals in a population to accomplish the change from any one particular stage of development to the next. There are, however, never more than two successive stages in the soil at the same time. The period of overlap of two stages in the population fluctuates widely (8 to. 41 days) from step to step in the development cycle. This is the result of the seasonal rise and fall in the soil temperature, and of diapause. The overlap, which is really a measure of range of developmental age among individuals, contracts to the minimum (8 days) as the population approaches maturity. This facilitates mating.

A summary is made of the literature dealing with the bionomics of Aphomia gularis (Zell.), a storage pest of almonds, walnuts, groundnuts and prunes, and to a lesser extent of rice and grain. Additional information is also given on its habits and occurrence in Britain.

An outline is given of its origins, introduction and establishment in various parts of the world. The evidence leaves little doubt that the species originated in south-east Asia, its occurrence elsewhere being, with few exceptions, confined to the major ports of western Europe and North America. Cases of the spread of the pest from south-east Asia can be traced to the export of infested goods from that area. Unless action is taken to prevent further dispersal and measures applied to wipe out the known centres of infestation, further establishment of this species can be expected in countries at present not affected.

It would appear that A. gularis is a subtropical and warm-temperate species, rarely found in the tropics and only able to maintain itself towards the northerly limits of its range, as in northern Britain and Sweden, in heated premises. In the cooler temperate regions, such as Britain, it cannot compete with such species as Ephestia elutella (Hb.) while in the tropics its ecological niche is filled by the very closely related species Corcyra cephalonica (Stnt.).

Its apparent absence from regions in the southern hemisphere, where conditions favourable for development exist, may be due to its presence not having yet been recognised.

The following description is based on material submitted for identification to the Commonwealth Institute of Entomology by the Government Entomologist, Coimbatore, S. India. The specimens were immediately recognisable as a new genus of the well-defined Mirid subfamily Bryocorinae tribe Odoniellini of which only 18 genera are known. Of these no less than 15 genera are represented in the British Museum (Nat. Hist.).

A number of predacious mites of the PHYTOSEIINAE (family LAELAPTIDAE), found in south-eastern England, mostly in association with fruit trees, are listed. Certain measurements and other characters that are of value in separating species are given. The species found were: Typhlodromus tiliae Oudm., T. cucumeris Oudm., T. tiliarum Oudm., T. rhenanus (Oudm.), T. finlandicus (Oudm.), T. umbraticus Chant, T. massei Nesbitt, T. vitis Oudm., T. soleiger (Ribaga), Phytoseius macropilis (Banks) and two Amblyseius spp. Of these, T. tiliae, T. finlandicus and P. macropilis are normally abundant on apple trees, but only T. tiliae remains abundant on commercially-grown apple trees. For each species a list of plants on which it has been found is given.

Details of the life-history of laboratory-reared mites are given and these, together with counts of field populations, show that three generations a year is normal.

Predacious phytoseiine mites feed on several species of TETRANYCHIDAE, including Metatetranychus ulmi (Koch) and Tetranychus telarius (L.); when M. ulmi is supplied in adequate numbers, adults of T. tiliae consume 3 mites per day, the nymphs 2 mites per day, on an average. It is thought that they also feed on plant tissue, since individuals survived in the laboratory for a considerable length of time in the absence of phytophagous mites, though eggs were not laid.

The host specificity of R. sanguineus (Latr.), 1806, sens, sir., and R. secundus Feldman-Muhsam, 1952, in Israel was studied.

It was found that 81 per cent, of R. sanguineus, sens, lat., found on the dog was R. sanguineus, sens, str., and 19 per cent. R. secundus, whereas on cattle, sheep and goats, 97, 94 and 91 per cent., respectively, were R. secundus and the remainder R. sanguineus, sens. str.

It is suggested that this situation exists also in other countries in the Near East, since R. secundus was very common in material from Algeria, Turkey, Yugoslavia and Iraq.

The problem of transmission of disease by R. sanguineus, sens, lat., should be re-investigated in the light of recent taxonomic data.

A portable wooden hut with two windows, on opposite sides, fitted with traps to catch mosquitos as they entered the hut was tried, with human bait, as a sampling device for mixed populations of Anopheles gambiae Giles and A. gambiae var. melas Theo. in and near Keneba village in the West Kiang district of the Gambia.

In eighteen night catches, between 24th August and 22nd September 1954, 5,901 females of this species and its variety were taken in the traps, five high catches (642 to 1,240 females) being made before 3rd September and thirteen low catches (27 to 306 females) between 6th and 22nd September. Possible causes for the different size of catches in the two periods are discussed with the conclusion that the differences in catch are not simply attributable to erratic functioning of the hut as a trap.

Four night catches of 186, 741, 797 and 1,240 females of A. gambiae and its variety are reported in detail. They reveal that mosquitos were taken virtually exclusively in the leeward trap if the wind persisted from one direction. But if the wind was variable, or conditions were calm, they were taken in both traps, entering the traps alternately as the wind shifted in the first instance and, perhaps simultaneously in calm. The results show that the use of only one trap could give misleading information about the numbers of mosquitos available outside the hut.

There is no doubt that, with wind, the mosquitos approached the host in the hut from down-wind. Since much of the catching occurred in darkness it would seem that the mosquitos were attracted by the emanations of the host and that their approach to the hut was not directed by a visual mechanism controlling up-wind flight.

Heavy rain for about seven hours did not prevent flight.

From hourly collections, made from dusk to dawn, a curve of activity for A. gambiae with its variety melas was obtained resembling the biting cycle reported elsewhere by other authors, the greatest activity occurring between midnight and 0600 hr. All the females (except two instances of incomplete fresh blood-meals) were unfed and, in a small sample, ovarian development was stage II or earlier. First entries occurred just before complete darkness in the evening and the last in good light about 30 minutes after sunrise.

No males of A. gambiae or var. melas were taken in the traps.

The numbers, all females, of other species taken were: 1 of A. funestus Giles, 1 of A. rufipes var. ingrami Edw., 59 of Taeniorhynchus (Mansonioides) spp., 30 of Culex nebulosus Theo., 19 of C. thalassius Theo., and 15 of unidentified species. Some Culicoides occurred in the traps and in the hut itself.

The hut, if modified to include a window trap in each of the four walls and a cowl over each window to keep out rain, appears to be a simple and convenient sampling device for certain types of investigations on the biology and behaviour of A. gambiae and its variety melas.

In working out a collection of Heterocera made by J. D. Bradley in the Solomon Islands, preparations were made of the genitalia of specimens of Spodoptera mauritia (Boisd.) and they were found to differ from those of African specimens, with which they were compared. Subsequent study of the material in the British Museum has shown that two species have been confused under the one name; S. mauritia, which occurs in Madagascar, Mauritius, the Comoro Islands and from India to the Pacific and which is known from continental Africa from only a single female taken at Lindi on the coast of Tanganyika, and the second species, S. triturata (Wlk.), which occurs throughout continental Africa, south of the Sahara Desert. As both species are of economic importance, it has been decided to describe their differences in a separate paper rather than include them in the faunistic paper dealing with the Heterocera of Eennell Island.

Trials are described with window-trap huts to test residual insecticides against vector mosquitos in Malaya. DDT, BHC and dieldrin were tested as wettable powders against Anopheles maculatus Theo., A. sundaicus (Ednw.) and Culex pipiens fatigans Wied. Some results were also obtained with Anopheles barbirostris Wulp, A. letifer Sandosham and A. umbrosus (Theo.), with species of Culex other than C. p. fatigans, with species of Mansonia, and with Aëdes albopictus (Skuse) and Aë. butleri Theo. All these mosquitos, except C. p. fatigans, are essentially outdoor species which enter houses only to bite, and which feed freely on animals as well as on man.

The validity of the window-trap hut method is discussed. It is concluded that although some mosquitos escape through the entrance louvres, and a treated hut is not quite the same as a treated and occupied room, the kills recorded are a good guide to the kills that may be expected in treated houses.

It is shown that there is usually some reduction, seldom sufficient to be important, in the number of mosquitos entering the huts in the first few weeks after treatment. This is probably due to air-borne particles of insecticide drifting through the louvres and acting on mosquitos waiting to enter.

The effect of treatment on the biting and resting behaviour of the mosquitos varied widely, being due to a combination of the properties of the particular insecticide and the habits and susceptibilities of the different species. Broadly speaking, all three insecticides tended to reduce the proportion of mosquitos obtaining a blood-meal, and the proportion remaining in the hut in the morning. But these tendencies might be modified or quite obscured by particular characteristics of the mosquito species or of the insecticide. Thus, for example, the percentage of A. maculatus obtaining a blood-meal was only reduced by 11 per cent, in the first month after treatment with DDT, compared with a 60 per cent, reduction in C. p. fatigans. It is suggested that perhaps A. maculatus did not touch the treated walls before biting and C. p. fatigans did. This difference applies to other species of Anopheles and Culex, and the average reductions in biting rate for the two genera were 19 and 62 per cent.

As another example, before treatment rather more than half of the C. p. fatigans, an indoor-resting species, remained in the hut; after treatment with DDT, which does not easily kill C. p. fatigans, over 80 per cent, were found in the window trap. By contrast, in the outdoor-resting species, A. maculatus, which is easily killed by DDT, over 90 per cent, were in the window trap before treatment, and fewer after treatment. The effect of DDT (and BHC) was to kill some of the A. maculatus before they could leave the hut, thereby reducing the proportion reaching the window trap to about 70 per cent. There was no reduction in the percentage of A. maculatus reaching the window trap with the slow-acting dieldrin.

Judged chiefly by the effect upon C. p. fatigans, DDT influenced mosquito behaviour mainly by its irritant effect, which persisted after the insecticide was no longer killing. The effect of BHC upon behaviour was due to a combination of irritance and rapid lethal action, but this did not last as long as the effect of DDT. Dieldrin has no irritant effect and it influenced behaviour only by lethal action for a short time while it was fresh; it continued to kill after it was no longer doing so rapidly enough to affect behaviour.

The most important finding was the wide range of susceptibility to the insecticides among the ten species or groups of species tested, and the fact that only two (Anopheles maculatus and A. umbrosus) seemed susceptible enough to be effectively controlled in practice by the fairly heavy doses used (200 mg. DDT, 40 mg. γ BHC or 40 mg. dieldrin) per sq. ft. With A. maculatus (and probably A. umbrosus) the 24-hour kill remained above 50 per cent, for about six months or more. With A. sundaicus (and probably A. barbirostris, A. letifer, Mansonia and Aë. butleri) the 24-hour kill fell below 50 per cent, in from one to four months. With C. p. fatigans (and probably Culex spp. and Aë. albopictus) the kill never reached 50 per cent, with DDT, and was only above 50 per cent, for one to two months with BHC and dieldrin.

A. maculatus was the most susceptible species to all three insecticides and C. p. fatigans the least; the latter was particularly insusceptible to DDT.

Heavier doses of dieldrin might be effective against the less susceptible species, and a dose of 100 mg. per sq. ft. has been found to remain effective for six months against Mansonia.

Except for the trial with light doses, dieldrin gave the best results against all species. When fresh it gave kills as high as or higher than those of BHC, and it remained effective longer than DDT or BHC.

Dieldrin and BHC when fresh gave complete or nearly complete kills of all species, but the rate of decline in the kills with time varied widely, and was quickest with the least susceptible species. With DDT, on the other hand, the rate of decline in the kill seemed to be roughly the same with all species, and it was the initial kill which varied. For the least susceptible group of species this was well under 50 per cent, in the first month, and was only 5 per cent, with one strain of C. p. fatigans; with the most susceptible it was between 80 and j 90 per cent.; it never reached 100 per cent.

Few mosquitos were found dead in treated huts, the great majority escaped into the window traps, especially with the slow-acting dieldrin. With A. maculatus in the first four months after treatment with DDT, only 25 per cent, of those dying in 24 hours were found dead in the hut. With BHC and dieldrin the corresponding figures were 14 and 2 per cent. The immediate kill (total found dead in the hut and window trap in the morning) was commonly less than half of the 24-hour kill. These results clearly show the importance of using window traps and holding mosquitos, if possible for 24 hours, if a reliable estimate of the effect of the insecticides is required.

Light doses (100 mg. DDT, 10 mg. γ BHC or 10 mg. dieldrin per sq. ft.) were tried against A. maculatus; DDT and BHC remained effective for three months, but dieldrin only for one month.

The performance of the insecticides at the higher doses can be characterised by saying that DDT is irritant and persistent, but not toxic enough except to the most susceptible of the species tested. BHC is irritant when fresh, though it kills at the same time; it is very toxic to all species, but does not remain effective long enough except against the most susceptible ones. Dieldrin is slower acting than DDT and BHC, but is non-irritant, very toxic, and remains effective longer; against less susceptible species higher doses than 40 mg. per sq. ft. will be needed. In countries where vector species rest indoors by day, dieldrin may prove particularly lasting because mosquitos will rest on treated surfaces for long periods.

The probable relation is discussed between the kills recorded in our windowtrap huts, and the degree of malaria control that may be expected when houses are sprayed. The kill necessary to control malaria depends to a large extent on how much contact the vector has with the insecticide. If its habits bring it into frequent contact with treated surfaces, a comparatively low kill on each occasion of contact may greatly reduce the population of the vector and suffice to control malaria. But if contact is infrequent, as may be expected with species of Anopheles such as those studied, which rest outdoors and feed only to a limited extent on man, then a high kill on each occasion of entry into treated houses is needed to control malaria, and there may be very little effect on the population of the vector. In these circumstances a 24-hour kill of 50 per cent., which is here considered the lower limit of effectiveness, seems a reasonable figure to adopt.

The mosquito factors which affect the degree of control achieved with residual insecticides can be thought of under three headings: (a) susceptibility to the insecticide, which largely determines the kill on each occasion of contact, and which we have shown may vary widely with different species, and is often quite low; (b) frequency of contact, which depends on habits, as these determine frequency of entry into treated shelters; (c) duration of contact on each occasion of entry, which is important when the insecticide deposit is no longer fresh, and depends mainly on whether the mosquito rests indoors by day and whether it is irritated by the insecticide.

The most important species of Pseudoeoccid transmitting the swollen-shoot virus disease in the Gold Coast is Pseudococcus njalensis Laing. It is almost always attended by ants of the genus Crematogaster. These ants usually build their nests in the cavities and galleries that have already been excavated by wood-boring insects in the dead branches of cacao trees. The only apparent advantage of this association to the ants is to imbibe the honeydew secreted from the anal orifice of the mealybugs; if this is allowed to accumulate it becomes a medium for bacteria and fungi which seems to kill the mealybugs eventually.

Attempts were made to break the link between the mealybugs and ants by spraying six cacao trees, not in contact with each other or any other trees, each with 3·5 litres of 0·2 per cent. DDT emulsion, twice at two-week intervals. All dead branches containing nests of ants were cut out, and a band of grease painted, on the trunk, one foot above the soil level to prevent the ants climbing up. Four weeks after the first application of spray, the population of mealybugs was brought down to 1·2 per cent, of its size before treatment. Unfortunately locating the ant nests, especially in the crevices and under the bark, is very difficult. Experiments were therefore carried out in which the trees were sprayed without previously eliminating all the sources of ants. The results were unsatisfactory, the size of the mealybug population, compared with that before treatment, after four, six and eight weeks was 16·9, 39·4 and 92·3 per cent., respectively, in the case of 0·2 per cent. DDT emulsion, and 11·6, 64·3 and 58·8 per cent., respectively, in the case of a treatment consisting of 3·5 litres per tree of a mixture of 0·2 per cent. DDT and 0·02 per cent, parathion active material. Cutting off the dead branches containing ant nests, followed by either painting the cut end of the dead branches with DDT (16·7 per cent, emulsion) or painting the trunk with a band of the same emulsion also gave poor results.

The unsatisfactory results obtained by spraying are attributed to the fact i that only the ants that happen to be exposed are affected by the application of ithe insecticide; the immature stages inside the nests are not touched.

The efficiency of dimefox when applied to the soil at 0·8 gm. active material per inch of tree girth was not increased when the ants had previously been partially eliminated.

Treated cacao trees could not in practice be completely isolated from forest jtrees and climbers containing nests of Crematogaster ants and these provide a continuous source of large populations. It is, therefore, considered that the icomplete elimination of ants is not possible.

It was also found that there is a relationship between the population of mealybugs and ants. Trees were usually free from mealybugs if no ants were found on them. The examination of isolated cacao trees with different sizes of dead branches containing nests of ants suggests that the number of mealybugs 38 in proportion to the size of the nests.

A correction is made to a paper by Nash (1955) in which Syntomosphyrum glossinae Wtstn., a Chalcid parasite of the pupae of tsetse flies, was wrongly recorded as parasitising the oöthecae of Periplaneta americana (L.). The parasite in question was almost certainly the very similar Tetrastichua hagenowii (Ratz.)which is a world-wide parasite of P. americana. There is no evidence that eitherparasite can attack both P. americana and Glossina palpalis (R.-D.).

Lucilia sericata (Mg.) has been bred successfully in small cages having a volume of about 1/60 cubic metre.

Female flies that contain ripe eggs, if given only sugar and water, will withhold their eggs for periods up to 24 days. There appears to be no decrease in the numbers of eggs laid or in the percentage of young larvae hatched, providing the delay in oviposition is not longer than ten days. If the meat is withheld for longer periods then there is a reduction in the daily egg production.

Further analysis of results obtained from a study of populations of Coccinellids on three plots of Vicia faba infested with Aphis fabae Scop, at Rothamsted during 1952 are presented. The distributions of coccinellid egg batches and older larvae (mostly 3rd and 4th instars) on bean stems of five arbitrary classes of aphid infestation, emphasise the differences in the manner of increase of the predator and prey populations considered here; at the same time, the results throw light on the oviposition behaviour of the adult Coccinellids.

The degree of correlation between coccinellid egg batches and Aphids on the plots (as indicated by regression analysis) varied inversely with the size of the aphid populations on the plots. On one plot, where Aphids were extremely numerous, this correlation was negligible; on the two other plots, where Aphids were less abundant, the correlation was more marked, being highest at the plot where the aphid numbers were lowest. The degree of correlation between egg batches and Aphids was determined apparently by the rate of increase of aphid numbers. It is pointed out that while these coccinellid populations increased additively (by immigration of ladybirds to the beans), populations of A. fabae increased multiplicatively (by reproduction) and at much higher rates.

During the early stages of the aphid infestations on the bean plots, when aphid-infested stems were comparatively few, there was no statistically significant association between coccinellid eggs and the presence of Aphids on the bean stems; but when all stems had become infested, egg batches tended to occur most frequently on the stems with the most Aphids. From these and other observations it is concluded that the female Coccinellids do not need the stimulus of the presence of Aphids before laying their eggs on beans, and that they concentrate on well-infested bean stems where they tend to stay and feed on the Aphids and probably oviposit on those stems. The distributions of older coccinellid larvae (3rd and 4th instars) indicate that they, too, tend to concentrate on well-infested stems.

The oviposition habits of the female Coccinellids are discussed in relation to the feeding problems of the newly-hatched larvae.