Two experiments were conducted simultaneously in 1962–63 at Chungai, in central Tanganyika, to test new equipment for disseminating insecticides from aircraft and a new insecticide, isobenzan (Telodrin), in the eradication of Glossina morsitans Westw. and G. pallidipes Aust. The new equipment consisted of an aerosol generator fitted to the exhaust of a Cessna 182E high-wing monoplane, and was used in both experiments. The isobenzan was compared with dieldrin, and the two insecticides were applied at rates inversely proportional to their toxicities to G. morsitans as previously determined in the laboratory.

Two blocks of woodland, each 11 sq. miles in area, were treated, one with a 12·3 per cent, solution of dieldrin at the rate of 0·0254 gal. per acre, giving a dosage of 0·5 oz. (14 g.) toxicant per acre, and the other with a 10 per cent, solution of isobenzan at the rate of 0·0124 gal. per acre, giving a dosage of 0·2 oz. (6 g.) per acre. The former block received eight treatments with dieldrin at approximately 3-week intervals, the fourth treatment being incomplete; the latter block received six treatments with isobenzan at intervals ranging from 20 to 45 days. The effects were assessed by means of fly-catches along fixed paths which continued for one year after treatments had ceased.

Both species of tsetse fly disappeared from the two blocks before the final treatments took place, and no more were caught until 11 months after spraying ended, when one example of G. morsitans was caught in each block; both were probably immigrants. None was found in the following month, and it is concluded that the flies were exterminated in both blocks. The fact that the blocks were unusually well isolated from sources of reinfestation probably contributed to the success of the operations.

The cost per sq. mile was £224 using dieldrin and £190 using isobenzan. These were the basic costs, independent of the locality in which spraying took place. Additional costs were incurred which would vary with local conditions; for the present experiments they were £34 and £30, respectively. It is considered that there is good scope for further reductions in costs, particularly with dieldrin, and that these might make dieldrin economically competitive with isobenzan. Owing to the low fly density in the block treated with isobenzan, the efficacy of this insecticide cannot be regarded as conclusively proved until further experiments have been carried out. It may then merit serious consideration as a toxicant for aerial spraying against tsetse flies.

Descriptions are given of 24-hour biting catches, made in and around Faza, a village on Pate Island, off the north-east coast of Kenya, East Africa.

Aedes pembaensis Theo. was the predominant mosquito in these catches but fair numbers of Aedes mombasaensis Mattingly were also taken; the biting cycles of these two are discussed. Six other species were taken in small numbers.

For Ae. pembaensis, biting cycles calculated on catches grouped for site, moon phases, neap tides and spring tides show that both moon and tide and light intensity influence the biting behaviour. Different but recurring patterns occur with various combinations of these factors.

For Ae. mombasaensis, the cycles have a very constant biphasic pattern. Catches grouped for moon phases, tides and catch sites, as for Ae. pembaensis, show that more biting females are taken at neap tides than at spring tides. Two patterns of behaviour occur, one associated with spring tides and the other with neap tides. An even level of biting activity occurs during the night with intense and prolonged moonlight; otherwise moon-phase cycles have little effect on biting behaviour.

It is felt that 24-hour biting catches can give useful information on behaviour patterns but it is desirable to have a long series of catches to analyse. For the purpose of calculating biting cycles, the results of catches showing similar modifications in behaviour should be treated separately.

Trials were carried out from 1958 to 1961 at Fassouri Plantation in Cyprus to assess the value of dimethoate (0,0-dimethyl S-methylcarbamoylmethyl phosphorodithioate) against California red scale, Aonidiella aurantii (Mask.), on orange, lemon and grapefruit trees. This pest has hitherto been controlled in Citrus plantations by annual programmes of sprays containing parathion, malathion or white mineral oil, singly or in combination.

Sprays were applied to Citrus trees at high pressure by means of hand-lance or motor-driven equipment in such a way that all parts of the tree were thoroughly wetted. Treatments were replicated in Latin square experimental designs. Results were assessed by counting the numbers of scales on samples of leaves, fruits, twigs and bark taken at various intervals after spraying and also, in some trials, before spraying. The earlier trials were concerned with the numbers of scales killed after single applications and the later trials with the level of control resulting from full seasonal spraying programmes.

A preliminary trial in 1958 showed that very high kill of the scale was produced by a spray containing 0·2 per cent, of dimethoate. In 1959 a range of concentrations from 0·01 to 0·15 per cent, was compared with the standard parathion/malathion mixture against the scale on orange and lemon trees. At 0·03 to 0·06 per cent., dimethoate gave a level of control equal to that of the standard spray and at higher concentrations it was superior to the standard. This superiority was confirmed in trials carried out during 1960 using dimethoate concentrations from 0·06 to 0·1 per cent., the observations being extended to include immature stages of the scale. A high kill of all developmental stages was produced by all the concentrations used, and at 0·08 to 0·1 per cent, dimethoate gave higher kills than the standard spraying mixture.

The effect of dimethoate against late-season infestations of fruits was next investigated, using mixtures with white mineral oil as well as dimethoate alone. Standard late-season sprays of mineral oil with or without parathion were included for comparison. Dimethoate alone at 0·08 per cent., or at 0·05 per cent, with 1·0 per cent, mineral oil, gave good control of the scale, on fruit. Both these sprays were superior, to the standard oil or parathion/oil mixtures. The results suggest that dimethoate may be especially valuable against scale on full-sized fruits, the control of which has hitherto been difficult.

During 1961, full seasonal spraying programmes with the standard compounds were compared with others using dimethoate at concentrations suggested by the earlier trials. The treated plots were large and the spraying methods approximated to normal estate practice. The final results were based on the level of infestation of fruits at harvest, and comparisons between standard and dimethoate programmes were made by means of Ebeling's corrected relative infestation index (C.R.I.). Values of the C.R.I, below 100 indicate that the degree of control achieved by the standard has been exceeded by the test spray. In the first trial, where early-season infestation was not heavy, the standard programme consisted of an initial parathion/malathion spray followed later by an application of white oil, and the test programmes comprised sprays of dimethoate at 0·04 or 0·06 per cent., both followed by white oil, or 0·08 or 0·1 per cent, without a second treatment. The C.R.I, values from these four programmes were 24, 22, 58 and 7, respectively, showing that all were better than the standard. The second trial was carried out in a plantation with a heavy early-season infestation of the scale; the same standard programme was compared with dimethoate sprays at 0·04 and 0·08 per cent., both followed by white oil. The C.R.I. values for these two programmes were 144 and 36, respectively. In the third trial single seasonal applications were compared, using a spray containing 1·5 per cent, white oil as the standard. The four test sprays comprised white oil 1·5 per cent, plus malathioii 0·05 per cent., white oil 1.0 per cent, plus dimethoate 0·04 per cent., dimethoate 0·04 per cent., and dimethoate 0·08 per cent. The respective C.R.I, values of the test sprays were 68, 38, 56 and 29.

In trials against other pests of Citrus trees dimethoate gave control of Coccus hesperidum L. at concentrations of 0·04 and 0·08 per cent. It was effective against Aceria sheldoni (Ewing) at concentrations of 0·08, 0·04 and 0·02 per cent., but was found to be ineffective against Phyllocoptruta olcivora (Ashm.).

During the trials a number of different formulations were used, including miscible oils and wettable powders. One of the miscible oils caused severe scorch marks on fruits; this was shown to be due to the formulating materials and not to dimethoate. A certain amount of leaf-drop was noticed in some trials, and this requires further investigation. Dimethoate is a comparatively safe compound from the point of view of hazards to operators, and toxic residues in sampled fruit have been found to lie within the limits that have been suggested as safe.

The principal conclusion from these trials was that dimethoate is a valuable addition to the limited range of insecticides available for use against Aonidiella aurantii. At fairly high concentrations (0·08%) it gives better control than spray mixtures in current use, and this suggests that it may be possible to devise programmes giving satisfactory control with fewer applications. At lower concentrations (0·04%) it is about as effective as present insecticides. Although dimethoate is both a contact and systemic insecticide, observations suggested that the former function was mainly involved in its action against this scale.

The inheritance of white eye in Lucilia cuprina (Wied.) was investigated by means of a series of back-crosses repeated through ten generations. Since an approximate 1:1 ratio of white-eyed: red-eyed flies was observed throughout, it is concluded that white eye is due to a single recessive gene (w). Earlier work establishing that w is on an autosome is discussed.

A test for the existence of a linkage group containing genes for white eye and dieldrin-susceptibility was then made. The exposure of back-cross progeny to discriminating levels of insecticide, and the pattern of mortality observed with scalar concentrations of dieldrin, indicated that genes for white eye (w) and dieldrin-susceptibility (rdi) are on different chromosomes.

The paper includes a description of studies on Anopheles gambiae Giles in the insecticide-free Umbugwe area of Tanganyika, and the results of dissections of A. gambiae and A. funestus Giles for sporozoite infections.

The densities of A. gambiae and A. funestus were assessed in houses by pyrethrum spray catches in 90 houses distributed in ten villages in the Umbugwe area between March 1960 and February 1962. A. gambiae was seasonally abundant, a sharp increase in density being associated with periods of rain. A. funestus type form was scarce in houses throughout the year, although a larval survey showed it to be the most common member of the complex of A. funestus in the area.

The average ratio of fed to gravid individuals of A. gambiae was 2·3:1 from June 1960 to November 1961, with peaks reaching 7:1 in March 1960 and 7·5:1 in January 1962. The peaks were attributed to rapid increase in production of adults during the limited period of the rainy seasons that followed very dry conditions.

Thirteen per cent, of recently fed individuals of A. gambiae were found in the exit window traps of experimental huts, having attempted to leave the huts on the night of feeding.

Two distinct types of huts are commonly found in the area, the tembe, which is large, divided into an average of five rooms by walls that are usually incomplete at the top, and accommodates large numbers of cattle as well as humans; and the banda, which is smaller, divided only into two rooms and seldom housing any domestic animals. Precipitin tests showed that whereas only 20 per cent, of individuals of A. gambiae taken in tembe had fed on man, in banda huts the percentage was 77. It was also shown that there was a dispersal of A. gambiae within a hut shortly after feeding. In a tembe. one-third of the individuals positive for man, and one-quarter of those positive for cattle had moved to rooms other than those in which they had fed.

A sporozoite rate of 1·47 per cent, was found among 3,746 specimens of A. gambiae dissected. Of the total number of 55 sporozoite-positive females of A. gambiae, 39 were found among the 424 individuals dissected in February 1962, thereby indicating unusually heavy transmission during the early part of 1962. A. funestus type form appeared to be of virtually no importance in malaria transmission in the Umbugwe area. Transmission by A. gambiae was highly seasonal, for although the percentage of gland-positive specimens was slightly higher in the long dry season than in the rainy season, the much larger mosquito population in the latter led to a higher number of infective bites in the rainy seasons. The highly seasonal transmission by A. gambiae was reflected in the results of examination of blood-films of people attending the Government dispensary.

A dipping technique for exposing large numbers of house-flies (Musca domestica L.) to measured doses of insecticide is described. It is suitable for selecting resistant populations and, while giving consistent results, is more rapid than other techniques used for this purpose.

Up to 2,000 flies of both sexes, less than 24 hr. old, are immersed for three minutes in 100 ml. of a 70 per cent, mixture of acetone and water containing the required concentration of insecticide, using a 9-cm. sintered glass Büchner funnel as the immersion chamber. The liquid is then removed by suction, the sides of the funnel are wiped with filter paper, and the flies are allowed to drain for three minutes; they are then transferred in small batches to plastic recovery chambers containing food. Mortality is recorded next day, and the survivors are released into breeding cages.

Experiments showed that immersion for three minutes in 70 per cent, acetone was virtually harmless to the flies and that the amount of insecticide deposited on individual flies was reasonably uniform (coefficient of variation about 20%). Batches of 2,000 flies, but not more, could be treated at one time.

When the dipping technique was compared with topical application of measured drops of insecticide, using a susceptible strain of house-flies and two other strains that were resistant to DDT and diazinon, dipping gave steeper log-probit regression lines than topical application, and the LD50's and resistance factors of the resistant strains were smaller. With flies resistant to DDT, dipping gave straight regression lines whereas topical application gave compound lines.

In both a susceptible and a diazinon-resistant strain of house-flies (Musca domestica L.), adults of both sexes were most susceptible to diazinon immediately after they had emerged from the pupae. Their resistance increased for two days after emergence and remained steady for the next three days, and the increase was proportionately larger in the resistant strain. Taking flies 2 ± 2 hr. old as a basis, the increases in resistance by the third day were 1−2½ and 3½−6½ times for susceptible and resistant males, respectively, and 2−5½ and 9−18½ for susceptible and resistant females. In both strains, the two sexes were equally susceptible to diazinon during the first day, but later the females were more resistant than the males. The resistance factor (ratio of LD50's of resistant and susceptible strains) was least immediately after emergence and increased with age; it was 15–30 in flies 2 ± 2 hr. old and 50–90 three days later. There was no direct relation between changes in resistance and changes in the weight of the flies on aging. The resistance factor to diazinon should be measured not earlier than three days after emergence to obtain reproducible values. Removal of the ovaries in flies less than one day old did not affect the increase in resistance with age.

The susceptibility of Musca domestica nebulo F. and M. d. vicina Macq. to topical applications of p,p' DDT, technical dieldrin and γ BHC was determined in Aligarh, India, and the effects of using acetone, Eisella oil or ethanol as solvents were also investigated. Of the three insecticides examined, BHC was found to be the most toxic to both forms of house-fly, and DDT the least toxic, though the latter would be effective for control purposes at concentrations exceeding 1 per cent. Solutions of the insecticides in ethanol were more toxic than solutions at the same concentration in Eisella oil, and the latter were slightly more toxic than solutions in acetone. The two forms of house-fly were about equally susceptible to DDT, but M. d. vicina was somewhat more susceptible to dieldrin and BHC than M. d. nebulo. Males of both forms were more susceptible than females to all three insecticides.

This paper records the ovarian condition of sylvan Culicine mosquitos caught resting and biting in Uganda. Information provided for the commoner species includes Sella's stages of blood-digestion, Christophers' ovarian stages (modified) for nullipars and pars, and sac stages for pars.

The distribution of Sella's stages among resting populations indicates that the several species differ widely in the resting sites chosen by blood-fed females. Among the species studied, Mansonia, fuscopennata (Theo.), M. africana (Edw.), M. uniformis (Theo.) and Culex annulioris Theo. are unusual in that many females probably rest outside forest after a blood-meal.

The ovarian stage in which nullipars and pars come to bite is often characteristic of a species. Nullipars and pars of one species may bite in the same ovarian stage, or in different stages. In species of Mansonia (Coquillettidia) nullipars are typically more advanced than pars, apparently because there is extensive use of food reserves which are exhausted before the first blood-meal; in such species ovarian development can also be discerned in pars between oviposition and a blood-meal, but it is extremely slight. In certain Aedes (subgenera Finlaya and Stegomyia), pars are typically more advanced than nullipars, and among those studied here it is only species of this type which to any extent show the pattern long regarded as normal in Anophelines. Wide variation in ovarian stages of biting nullipars of Aedes raises the possibility that limited autogeny may exist in some of the species studied.

Observations were carried out, chiefly between March and June 1962, in an area centred on Bolgatanga in northern Ghana in the hope of finding out how Simulium damnosum Theo. survives the ‘ dry season ’. The term ‘ dry season ’ is used here for the unfavourable season when running water in the rivers is evidently absent or is insufficient or perhaps unsuitable for the development of any, or more than a negligible number of, individuals of S. damnosum from egg to adult. The incidence of this season varies from year to year and does not correspond to the season without rain. It may usefully be divided into reduced-flow, main dry, pre-flow and early-flow periods.

Search was made for larvae and pupae before rivers began to flow, without result. Observations were made on the numbers of flies biting, general appearance, wing measurement, antennal colour, mid-gut, fat-body, ovaries and parasites, chiefly at the beginning of the rainy season, in an attempt to trace the history of the species during the preceding weeks. Flies appeared before the rivers began to flow but did not give evidence of having lived for months. On the days when two rivers began to flow, considerable numbers of flies were caught, and most of those dissected proved to be parous and had evidently been in the adult stage for some days. A nematode parasite of the Malpighian tubes was found only in April, with one exception up to 10th May, and was probably associated with the dry-season biology of the host.

The possibility that S. damnosum migrates from permanent foci, or exists near its breeding places in one or more stages at different times, is discussed. Migration from afar seems very unlikely. The species survives in the adult stage for at least part of the ‘ dry season ’, but more information is required to complete the picture.

Survival is apparently not dependent to a significant extent on perennial breeding locally, and there is, therefore, no easy solution to the problem of control. Operational research might show if there is a carry-over generation and if control after the rains and towards the end of the breeding season would prevent it from developing. There is some indication that the risk of infection with Onchocerca tends to be low at the end of the ‘ dry season ’, which is a time when control would be particularly difficult.

The results of observations on wing measurements, antennal colour and the peritrophic membrane are given in an Appendix.

Allatectomy of newly emerged house-flies (Musca domestica L.) prevents maturation of the ovaries, and experiments were done to find whether this affected the sensitivity of the flies to insecticides.

The apparatus and operational procedure for removing the complex comprising the corpus allatum and corpora cardiaca are described.

Females of a diazinon-resistant strain (SKA) were allatectomised less than 12 hr. after emergence, and when three days old their sensitivity to diazinon, dieldrin and pyrethrum extract was tested by topical application. Approximately equal numbers of intact females were tested simultaneously, and untreated controls of both kinds of flies were included in every test. The failure of the ovaries to mature in allatectomised flies was confirmed by examination of a majority of the operated flies that lived for four days, and normal development of the ovaries of intact flies was similarly confirmed.

The results showed that there was no difference in sensitivity between allatectomised and normal flies. It is concluded that allatectomy does not affect susceptibility to the insecticides used. Since earlier work had shown that a 9- to 18-fold increase in resistance to diazinon takes place during the period from emergence to the third day afterwards, it is further concluded that this increase is unaffected by allatectomy and the failure of the ovaries to mature.

Bagged maize stored in large stacks on the Kenya coast is subject to reinvasion by stored-products pests following successful fumigation with methylbromide, and treatment of the stack surfaces with insecticidal sprays has been unsatisfactory in preventing reinfestation.

Further tests are described of different methods of applying insecticides to small experimental stacks, each composed of ten 200-lb. bags of maize, exposed within a warehouse, after fumigation, to reinfestation from a large bulk of maize heavily infested with Tribolium castaneum (Hbst.). Populations of adults and pre-adult stages of T. castaneum were recorded from samples of grain obtained by a sampling spear from each bag at intervals.

In the first of two experiments, single applications of malathion in a spray at 100 and 150 mg. per sq. ft. to the entire external surface of each bag during the construction of the stack did not give good control of reinfestation in the six months after fumigation, but were not significantly worse, respectively, than a dust containing 1 per cent. lindane similarly applied to the upper surface of each bag and, four weeks later, to the outside of the stack at the rates of 3 and 6 oz. per bag. When the malathion treatments were followed by 24, 12 or 8 applications, at intervals of 1, 2 or 3 weeks, respectively, of sprays containing malathion to give 100 and 150 mg. per sq. ft. over the whole external surface of the stack, good control of T. castaneum was obtained with all treatments.

Mean residues of malathion in the outer inch of maize in the bags six months after the start of the experiment varied from 3 to about 19 p.p.m., according to treatment, suggesting that little penetrated through the bag wall and that after thorough mixing of the grain before use the malathion content would not exceed the permissible limit.

In the second experiment, an admixture of malathion dust with the maize before bagging to give 8 p.p.m. active ingredient gave almost complete control of larvae and pupae, and application of 3 oz. 0·5 per cent, malathion dust, or a spray affording 400 mg. malathion per sq. ft., to the upper surface of each bag during stacking gave highly significant reductions in numbers of T. castaneum in the pre-adult stages. The last of these treatments, followed, after fumigation, by malathion applied in a spray at 150 mg. per sq. ft. every three weeks to the external surfaces of the stack, is recommended for large-scale trial.

In view of the promising results obtained with insecticidal smokes against the common furniture beetle, Anobium punctatum (Deg.), a field trial was conducted to study the effect of deposits from lindane/dieldrin smoke on adults of the death-watch beetle, Xestobium rufovillosum (Deg.), emerging from infested roof timbers of a college chapel in Cambridge, England, in 1963. Quantitative determination of the insecticidal deposits was made and methods were devised for studying the effect of the treatment on the numbers of beetles emerging subsequently in order to assess the long-term value of annual treatments for reducing the population. Beetles collected from an untreated building were used for comparison.

Fewer males were found than in preceding years and it is postulated that some may have been prevented by the insecticides from emerging. The normal sequence of beetles dropping from roofing timbers—males predominating first, then females—did not occur. Both sexes dropped nearly uniformly, a fact suggesting, particularly in the case of the females, that the insecticides had caused them to drop before their normal time. Mating was considerably reduced. Examination of females showed no signs of their having laid eggs, this being attributed to the action of the insecticide deposits. Beetles obtained from an untreated building and placed on wood surfaces which had been exposed to the smoke treatment were affected but laid some eggs, of which a small proportion hatched; larval boring into such surfaces was, however, completely prevented.

From these results it was inferred that the deposits from the treatment with lindane/dieldrin smoke prevented emerging adults from causing reinfestation of the roofing timbers and it is argued that, provided that this can be achieved annually by a succession of treatments, the beetle population should ultimately be eliminated or, at least, greatly reduced. Other possible effects which might result from annual smoke treatments are accumulation of the insecticide deposits with a consequent increase in toxicity of the timber surfaces and sorption of the deposits by the timber to a point where they might exert a toxic effect on larvae or pupae beneath the surface.

Five trials were conducted in Kenya in 1962–63 to investigate the effectiveness of a 0·2 per cent, w/w pyrethrum dust, synergised with 5 parts piperonyl butoxide to 1 part pyrethrins, against the stalk borer Busseola fusca (Fuller) by applying it to maize at 20 lb. per acre. In three trials, it was compared with a dust containing 5 per cent. DDT applied at the same rate. Pyrethrum showed an efficiency in controlling B. fusca equal to that of DDT. Both insecticides produced reductions of more than 50 per cent, in the incidence of infestation, and in the final trial there was a significant reduction of over 40 per cent, in the number of cobs damaged by stalk-borer larvae; at the same time, an increase in yield was apparent. Further work is required in order to determine the lowest effective concentration of pyrethrins in the dust, which is likely to be less than 0·2 per cent., and to indicate the best time of application.

Gravid females of the red locust, Nomadacris septemlasciata (Serv.), were initially influenced by the microclimate in the selection of oviposition sites, and a laboratory investigation showed that females preferred to lay in well-illuminated areas with a soil temperature between 43 and 37°C., and avoided areas of very high surface humidity. Final acceptance of the soil took place only after the locust had probed with its ovipositor, and experiments demonstrated that particle size, compaction, moisture content and chemical composition of the soil all influenced the choice of oviposition site, while pH did not.

The larvae of two species of swift moths, Hepialus humuli (L.) and H. lupulinus (L.), live in the soil and are common under grass in the British Isles, but do not usually have any obvious effect upon the grass. The larvae of H. humuli are polyphagous and when they occur under agricultural or horticultural crops they may cause damage by feeding on the roots. Lettuce, strawberry and chrysanthemum are most frequently damaged.

The adults, which are described, fly at dusk, the female seeking the male. The female lays between 200 and 1,600 eggs, with a mean of about 600, over a period of four days. Sweep-net catches showed the ratio of the sexes to be approximately equal, and mercury-vapour light-trap catches showed that the flight period in southern England is principally in June.

In the laboratory, never less than 80 per cent, of eggs hatched, even in dry air, and the shortest incubation period, between 11 and 24 (mean 18) days, was at 20°C. At 5°C. they did not hatch, but remained viable for at least six months.

The larvae were reared in small cells drilled in blocks of plaster of Paris, which were stood in a tray of water, or in vials containing a layer of moist plaster of Paris, and were fed on pieces of carrot.

The growth of the larval head capsules was geometric and allowed 12 instars to be distinguished at 15°C.; at higher temperatures there may be more.

The growth curve of the larvae was S-shaped with a very distinct fall in weight in each instar for several days before moulting. The daily intake of food was weighed, and the conversion rate shown to be small, ranging from 0·00058 to 0·04, as compared with 0·205 to 0·49 recorded for other Lepidopterous larvae.

At 5°C., larvae did not develop beyond the second instar, nor beyond the fourth at 10°C. The optimum temperature for their development (mean period 224 days) was 15°C., although mortality was high. They developed at 20°C. more rapidly (mean 197 days) but more died. Larvae kept in a container sunk in the soil outdoors developed more slowly than those at 15°C. and had only reached the eighth instar by February (239 days after hatching). This evidence, together with that from an artificially infested plot in the field and general field observations, suggests that the usual life-cycle lasts two years.

Experimental evidence is given to show that only two larval instars occur during the development of Meligethes aeneus (F.) (Coleoptera, Nitidulidae) under laboratory conditions in Scotland, based on the facts that there is only one moult and only one increase in head-capsule width. The egg, first- and second-instar larvae, prepupa and pupa of M. aeneus are described and compared with the corresponding stages of M. viridescens (F.). The larvae of the two species are differentiated for the first time by the arrangement of dorsal cuticular tubercles, the size and form of the urogomphi and the width of the head capsules.

The majority of sorghum in Northern Nigeria is stored unthreshed in farmers' granaries made of dried mud or plant materials such as grass matting and cereal stems. During the course of survey work and insecticide trials in 1959–61 it was possible to examine many samples of unthreshed sorghum from granaries throughout Northern Nigeria. It was found that the distribution of insect species within the Region is not uniform. Sitophilus oryzae (L.) and Sitotroga cerealella (Ol.) are the major pests. Sitophilus zeamais Motsch. was found only in the southernmost area. Heavy infestations of Rhyzopertha dominica (F.), Lasioderma serricorne (F.), Cryptolestes ugandae Steel & Howe, Oryzaephilus mercator (Fauv.), Tribolium castaneum (Hbst.) and T. confusum Duv. usually occur. Attagenus gloriosae (F.), Ahasverus advena (Waltl), Palorus ficicola (Woll.), P. ratzeburgi (Wissm.) and P. subdcpressus (Woll.) are occasionally important.

R. dominica, Bruchidius sp., G. ugandae, Planolestes cornutus (Grouv.), S. oryzae, Brachypeplus sp., T. castaneum, Sitotroga cerealella (all of which occurred on the standing crop), L. serricorne, Typhaea stercorea (L.), O. mercator, Palorus spp. and Tribolium confusum were found in sorghum sampled before storage. Insects from infested stores were found to infest sorghum growing nearby. Prestorage infestation alone can result in subsequent heavy populations of insects in the store. Cross-infestation between granaries almost certainly occurs.

Under Samaru conditions, where sorghum is harvested in November–December, insect populations remain at a low level in granary-stored unthreshed sorghum until after June, when the moisture content rises in the rainy season. In sorghum stored for nine months in provincial trials, more damage occurred during the last three months than during the previous six months of storage.

Sorghum heads stored in granaries made of plant materials such as grass matting and cereal stems are more severely damaged by insects than those stored in dried-mud granaries. This is probably due to a higher rate of immigration in the former.

In threshed grain stored in a mud granary, moisture content and damage by the most abundant insect, Sitophilus oryzae, decreased with depth. The insect population rose to a peak in November, two months after the maximum moisture content. The numbers fell rapidly during the following dry season. The annual cycle of insect damage was also assessed by taking fortnightly samples of threshed grain from a local market. S. oryzae was the most numerous insect, but even this species was uncommon from December to May, during the dry season.

The three-dimensional pattern of dispersion of five species of grain-infesting beetles was analysed in a 12-in.-cube bulk of wheat, kept in darkness at 25°C. and 14 per cent, moisture content, and built up of four layers, each of 16 three-in.-cube bags of Terylene net; the net retained the grain but allowed passage of insects. One hundred adults were placed in the centre of the upper surface of the topmost layer and after two weeks the entire bulk was broken down and the number of insects in each bag was determined. Five replicates were used for each species. More recoveries of Sitophilus granarius (L.), Tribolium castaneum (Hbst.), Rhyzopertha dominica (F.) and Cryptolestes ferrugineus (Steph.) were made from the top quarter of the bulk than from any of the lower layers. The vertical dispersion of Oryzaephilus surinamensis (L.) was homogeneous. Lateral dispersion was homogeneous with T. castaneum, R. dominica and C. ferrugineus, but with S. granarius and O. surinamensis relatively more recoveries were made from the corner units of the bulk than from elsewhere. The dispersion pattern assumed by a species in a uniform bulk is affected by locomotory activity, boundary response and intragroup stimulation. Thigmotaxis was demonstrated in S. granarius, persisting at relatively high densities; thus, following dispersal, individuals collect along habitat boundaries and then move up them as a result of intragroup stimulation. The positive response to boundaries shown by individuals of T. castaneum in isolation breaks down in crowded conditions, producing movement back into the bulk of grain and thus homogeneous lateral dispersion. O. surinamensis shows a positive response to habitat boundaries but no movement up them. Dispersal is slow in R. dominica, but the ultimate pattern of vertical dispersion is homogeneous. C. ferrugineus shows a tendency to form clumps, even in conditions free from marked physical gradients. The data are discussed in relation to ecology of the species, low-density populations, aggregation behaviour and a dynamic concept of population dispersion.