A light-weight spear has been developed to take samples of unlimited size from bulks of cereal grains. The spear is screwed on to lengths of narrow metal pipe, which are used to insert it into a bulk of grain. The sample is sucked out of the bulk through the pipe by the air stream of a domestic vacuum cleaner. This method largely overcomes the disadvantages inherent in other types of spear. The spear can be inserted into a bulk in any direction, even upwards. When sampling downwards, samples can be obtained from as deep as 9 m. Samples can be taken from directly against the walls of a store and as close as 1 cm. to the floor.

In the course of recent investigations of the association of ant colonies with subterranean aphids, it became necessary to develop a suitable method of culturing the latter in order to facilitate experimental work.

Parts of a district situated in the tropical rain-forest of the Belgian Congo in which onchocerciasis was generally endemic, were investigated to identify the vector and to ascertain its phoretic host, breeding places, habits, etc.

The vector found biting human bait (usually less than 1 metre from the ground) was Simulium neavei Eoub. Dissection of adult females revealed developmental forms of Onchocerca volvulus.

Larvae and pupae were found on crabs of the genus Potamon in fast-flowing stretches of the numerous small streams flowing into the Eiver Congo. The infested streams in the main focus of onchocerciasis were treated by DDT emulsion at a concentration of one part per 500,000 for 30 minutes, every ten days for ten applications.

The good results observed initially were partly offset by the reappearance six months later of small numbers of Simuliids apparently developing from eggs laid by females entering the disinfested area from without.

Cross-mating experiments are reported between adults of the typical, darkclubbed form of Syntomosphyrum glossinae Wtstn., obtained from pupae of Glossina pallidipes Aust. from Kenya and of G. palpalis (R.-D.) and G. morsitans submorsitans Newst. from Northern Nigeria, and the ‘white-clubbed’ form of Syntomosphyrum, originally reported by Waterston from Nyasaland, obtained from pupae of G. morsitans Westw. from Tanganyika.

Fertilised eggs, giving rise to offspring of both sexes, were produced only by crosses between the strains of typical S. glossinae. Crosses in either direction between the ‘white-clubbed’ form and typical S. glossinae resulted in all-male progeny, presumed to develop from unfertilised eggs. It is considered that this is evidence of a physiological barrier to reproduction between the two forms. The ‘white-clubbed’ form is therefore regarded as a distinct species.

The object of this paper is to give taxonomic consideration and status to the ‘white-clubbed’ parasite of tsetse flies that was studied by Dr. D. S. Saunders in the foregoing paper. Since the biological evidence in conjunction with the distribution suggests that it is specifically distinct from Syntomosphyrum glossinae Waterston it is described as a new species, although the morphological separation is weak.

The life-cycle and immature stages of the Eulophid, Syntomosphyrum albiclavus Kerrich, a pupal parasite of tsetse flies (Glossina), are described. Measurements of the widths of the cephalic skeletons of larvae of all ages fell into four well separated groups, and confirmation that these represented the existence of four larval instars was given by Dyar's law, the application of which in this instance is not invalidated by the disproportionate length of the fourth instar or the effects of overcrowding, which do not become apparent before that stage.

A convenient and accurate measure of size is necessary for studies on the physiology of tsetse flies (Glossina). Residual dry weight (i.e., dry weight minus fat) is adequate as regards the larva, pupa and teneral fly, but cannot be used for flies that have taken their first blood-meals. A linear measurement provides an index of size that remains constant throughout the life of the fly, but the use of the length of the middle part of the fourth longitudinal vein in the wing has certain disadvantages, and a measure of surface was thought preferable. That finally adopted was a function of the dorsal thoracic surface; details of it are given and a close correlation was demonstrated between it and residual dry weight for females of G. morsitans Westw. bred in the laboratory from puparia maintained at 28°C.

The relation between this measure and the weight of the newly emerged fly is affected by the temperature at which the pupal stage is passed; above approximately 25°C. the surface area of flies of any given residual dry weight decreases quite sharply, so that puparia of a given weight produce smaller but relatively heavier flies at 30 than at 25°C. This holds good also for G. pallidipes Aust. and G. swynnertoni Aust. The dependence of surface area on the temperature at which development has taken place unavoidably complicates the use of this measurement in work on transpiration; it also limits the application of the theory that stresses acting on the pregnant fly may be gauged from the size of flies two months later, and suggests that where the mean temperature of pupal sites is likely to exceed 26°C, any inference should be based on puparial rather than adult size.

The extent of pigmentation of the abdominal bands of Glossina pallidipes Aust. was studied in specimens emerging in the laboratory from puparia maintained at constant temperature and a range of relative humidities and derived from females collected in the field at Shinyanga, Tanganyika. An index of pigmentation was obtained by regarding the bands as consisting in all of 15 zones, each carrying a score of 1.

The index was closely dependent on the humidity experienced during pupal development, almost full pigmentation being produced in saturated air and almost complete suppression of it in dry air; it is inversely related to size and little affected in nearly saturated air by temperatures below 28°C. during pupal development, although still higher temperatures reduce it quite sharply.

In flies collected in the field throughout the year, most of those taken towards the end of the rains (March and April) fell into the darker pigment categories; as the dry season advanced the distribution tended to become bimodal, with the darker mode relatively steady at an index of about 12 and the paler mode moving progressively down the scale, reaching an index of about 5 in October and then disappearing. December and January collections (rainy season) were again unimodal, resembling those of March. It is suggested that these changes reflect the use of two types of larviposition site, the darker mode representing pupae from the evergreen thickets along the drainage lines and the lighter mode those from semi-deciduous thickets on the eluvial slopes, and that during the dry season the soil-space humidity in the latter sites may reach levels inimical to survival.

The depth to which larvae of G. pallidipes burrow appears to be related to the temperature, and perhaps the humidity, obtaining at the time of larviposition.

The possible effect of humidity on coloration was tested in four other species of tsetse fly, namely G. swynnertoni Aust., G. morsitans Westw., G. palpalis fuscipes Newst. and G. longipennis Corti. Only G. swynnertoni gave indications of some slight effect.

Earlier work in Nyanza Province, Kenya, had led to the conclusion that riverine communities of Glossina palpalis fuscipes Newst. were permanently hungrier than lake-side communities. This was based on analysis of catches (percentage of teneral flies in the whole catch, and percentage of females among non-teneral flies) and on a behaviour criterion (percentage of non-teneral males caught on the party), but subsequent observations in the two types of habitat, in which a different area was used to represent the lake side, showed the lake-side flies to be slightly smaller than the riverine flies and not to differ in fat content, and a further investigation was accordingly made.

Puparia were collected on three dates between January and March 1958 and their surface area measured. A measure of the thoracic surface and fat content was taken of all adults emerging. Non-teneral flies were collected, for comparison, about two weeks later than the puparia.

In all samples, the riverine puparia were significantly larger than those from the lake side, and flies that emerged showed a comparable difference in size, but not in fat content corrected for size. Non-teneral riverine flies were slightly larger than lake-side ones but the difference was less than half that shown by the puparia and the teneral flies obtained from them. Evidence is adduced to show that this discrepancy is due to elimination of the lower size-groups in the lake-side community. Estimates of fat content of non-teneral flies support the conclusion that the lake-side community was under stress as compared with the riverine one, and analysis of available fly-round data from the lake side also suggests stress conditions.

The lake-side results are at variance with those from the site used in the earlier work. Available meteorological data show that the average temperature during February and March is 2°C. higher at the present site than at the former; such difference may directly cause the stresses that appear to be operating in the present instance.

The results suggest that the percentage of non-teneral males caught on the party has a real meaning in terms of nutritional status, and that it is also a very responsive indicator, but that the proportions of tenerals and of female nontenerals may not be simply related to the nutritional state as had hitherto been supposed.

Reasons for the large day-to-day variations in the numbers of nymphs and adults of the Coreid bug, Pseudotheraptus wayi Brown, taken in the crowns of coconut palms in Zanzibar by the searching technique over a period of six years were investigated in 1957–58. Data are given showing the results of searches (each occupying less than 20 minutes) carried out on two young palms at different times of day and night and in different weather conditions. For each successive 2-hour period between 6.0 a.m. and 2.0 p.m. (in the case of nymphs) or 4.0 p.m. (in the case of adults) the proportion of searches in which P. wayi was observed was significantly greater on dull days than on bright, sunny days, but there were no such differences that were significant in respect of any other 2-hour period for which comparisons could be made, except 4.0 to 6.0 p.m., when the proportion of searches disclosing nymphs was greater on bright days. During the hours of bright sunlight, the adults and nymphs hide in crevices and in the debris found in the leaf axils, but both remain in the open from dusk to dawn. During heavy rain they take shelter in dry crevices and niches.

It is pointed out that this behaviour must be taken into consideration if the searching technique is used to assess the result of an application of insecticide.

Dysdercus superstitiosus (F.) is capable of utilising a much wider range of food-plants than was previously supposed, and is able to complete a full breeding cycle on the several diets other than cotton. This has been demonstrated by rearing and breeding adults and nymphs in the laboratory, in Uganda, on selected food material, e.g., Sorghum vulgare, Pennisetum typhoides, Zea mays and Vigna unguiculata, immature seeds of each being used. Diets other than mature cotton seeds are less suitable for breeding, as judged by fecundity of females and development of nymphs, but the influence of nymphal diet on adult fecundity is not consistent. Females reared on sorghum are smaller than those reared on cotton seed and they may lay fewer eggs.

There is an indication that varieties of sorghum may differ in their suitability for producing maximum fecundity and fertility. Those varieties to which the stainer exhibits a predisposition in the field may not be the most suitable.

When adults are fed on cotton bolls of ages 1–10 weeks, the fecundity of females, and the rate of egg-production increase with increasing age of boll, and the preoviposition period decreases.

Stored cotton seed may be less suitable for egg-production than freshly harvested seed cotton.

Extraction of oil from cotton seed with petroleum ether does not wholly impair the fecundity rate of females fed on the seed, but may reduce the fertility and nymphal survival in the first instar.

A diet of cotton seed, heat treated at 150–190°C. for one hour, reduces the fecundity in the females, possibly by destroying the accessary food factors. Nymphal survival is possible on such treated seed, but not when thus heated for two hours at or above 150°C.

Some 246 species of mosquitos are known to occur in Uganda. Of these, 92 (37·4%) have been recorded as breeding in swamps, and the present paper brings together published data and the results of work on the collection and identification of larvae from habitats in a wide variety of types of swamp between 1955 and 1958. Notes on the occurrence and habitats of the larvae are given under each species.

In the present work, only 58 species were found breeding, but these included six new swamp records. Only 26 species appear to breed exclusively in swamps. In addition to the species identified, larvae representing some 14 unrecognised and probably undescribed species were collected. The majority of the swampbreeding species are Culicines, the Anophelines comprising only 21·7 per cent.

The swamp environment in Uganda, with respect to the breeding of mosquitos, is extremely varied. Some possible classifications of the many and various swamps found in the country are given. The distribution of certain species of mosquitos is more or less limited to certain types of swamp This is briefly discussed and examples are given. There is also a definite zonal distribution of some species within a swamp, e.g., Culex (Culex) grahami Theo., C. (C.) guiarti Blanch, and Ficalbia (Ficalbia) malfeyti Newst. occur only in peripheral zones. In general, the interior of the large swamps is unfavourable to the breeding of Anophelines, but Culicines are very abundant there.

Breeding of mosquitos is profoundly affected when swamps are altered by human interference. In certain cases this has resulted in increased production of Anophelines, with the consequent aggravation of the malaria situation.

It is concluded that, from the point of view of human disease, the swamps of Uganda (especially in their natural untouched state) are not as dangerous as previously thought.

The paper deals with two members of the Simulium neavi Eoub. complex found at Amani in the forest of the Usambara Mountains, Tanganyika Territory. They have already been described but are at present termed the Amani unhanded and banded forms, scientific names being witheld until the taxonomy of the whole complex is better understood.

Some methods of study are described.

Pupae of the unhanded form constituted a large proportion of the total number found on the common local crab, Potamon (Potamonautes) lirrangensis Eathbun, which amounted to roughly one pupa to seven crabs. This form predominated in the lower and more open parts of the drainage system examined. The adults were not seen to bite man. This and the banded form were found at a much lower altitude than any hitherto reported for the complex, probably owing to the particular nature of the local climate. Neither form was abundant.

The larvae and pupae of the banded form, like the unhanded one, live externally on the crabs. The variability of wing size of the banded form is described. Observations were made on the internal anatomy of the adult with special reference to the ovaries in which the follicular relics are usually very large, indicating that the female probably bites soon after laying eggs. As in S. damnosum Theo. there is an easily recognisable class of old flies with clear Malpighian tubes. Nulliparous and parous flies have rather different biting cycles. The banded form bit man readily but only about five were taken per hour in good catching places and far fewer elsewhere.

Human onchorcerciasis has been reported from Amani but is not known to do any harm there. The existence of the disease in the presence of rather small numbers of Simuliids is of interest in relation to a belief that S. neavei is a more efficient vector of onchocerciasis than is S. damnosum. Out of 359 banded females dissected 41·2 per cent, were parous and 12·8 per cent, of these were infected with nematodes, at least some of which were not Onchocerca volvulus.

Taxonomy, the ovarioles and the Simuliid-crab association are discussed. The formation of the large follicular relics of the banded form is considered in relation to the history of the follicular epithelium before and after ovulation. The crab association is regarded as commensalism-, and it is suggested that the crabs benefit the Simuliids by carrying them a short distance to a suitable current and that they thus enable them to live in fast rivers and so extend their range.

Baris granulipennis (Tourn.), originally described from Egypt, has become a serious pest of water-melons (Colocynthis citrullus) in Israel, where hitherto its presence has not been recorded, and has been found on melons and cucumbers, but not on vegetable marrow or squash. Economically serious damage results from girdling of the stem end of the fruit by the female, prior to oviposition.

The incubation period of the egg, on average, is 3·7 days at 30° and 6 days at 22°C. The larvae feed on developing seeds inside the fruit and pupate there. The total development period occupied from 25 to 42 days in the summer, when the temperature fluctuated between 24 and 29°C. The cocoon is made by glueing minute pellets together. These consist of dried pulp from the fruit mixed with a secretion which is ‘milked’ from the anus, and is believed to be produced in the Malpighian tubules. The preoviposition period is about four days, and the length of adult life is several weeks. There are three generations a year, and four if weather permits. Adults overwinter in hibernation. It is suggested that the increase of cucurbit cultivation, and the lengthening, by irrigation, of the cropping period, may have permitted an increase in the weevil population to pest proportions.

An account is given of the biology of the phytophagous Chalcidoid, Eurytoma sp. near howardi D.T., and of its bionomics in Mauritius. The insect was introduced into Mauritius from Trinidad to supplement the activity of the Galerucid beetle, Schematiza cordiae Barber, in the control of Cordia macrostachya, an introduced shrub, indigenous to Central America and the Caribbean Islands. The egg and larval stages are described and figured.

The entire development of Eurytoma takes place within the fruit of Cordia, which contains usually only one seed when mature and has a very hard and stony endocarp and a soft and fleshy mesocarp. Oviposition by Eurytoma takes place before the endocarp hardens; the egg is laid inside the seed and the larva feeds on the nucellus and endosperm and later on the developing cotyledons, until the entire seed, except for the seed coat, has been consumed. The attacked fruit continues to grow and its endocarp hardens, but final swelling of the mesocarp does not take place. Although more than one egg may be laid in a seed, only a small percentage of the many seeds dissected contained two mature larvae or two pupae. Pupation takes place within the seed, from which the adult escapes by boring a round hole through the endocarp.

Development from egg to adult occupies about 23 days in the hot season (November–April), but in the cooler months (June–September) the fully-grown larva enters a resting period, which may last several months.

Adults kept in jars, and provided with honey in water, at a mean temperature of 25°C, lived for an average of 11 days (males) or 20 days (females). A newly emerged female contains about 12 fully developed eggs. Virgin females oviposit readily, all the progeny being male.

Releases of adults from material received from Trinidad were made between October 1949 and May 1950; by the latter date it was evident that the insect had become established at the first liberation site and further colonies were established in various parts of the island from material collected there. Estimates of infestation of samples of fruit collected in October 1951 at or near 22 liberation points ranged from 16 to 81 per cent, (mean, 53·7). The latest island-wide estimates, in February, June and October 1953, showed averages of 64·7, 40·8 and 48·7 per cent, infestation, respectively.

Infestation of fruits is highest (70–80%) in the hot season, when fruit production is at its maximum, and lowest in the cooler months, when, although the rate of fruiting declines, the life-cycle of Eurytoma is prolonged.

Schematiza cordiae had achieved a high degree of control of Cordia before Eurytoma became widely established, and it is difficult to assess the contribution made by the latter, which has no known specific natural enemies in Mauritius; but the two unrelated insects acting jointly are likely to stabilise control more effectively than would one insect alone.

In an appendix, P. 0. Wiehe gives notes on the present ecological status of Cordia in Mauritius and contrasts it with that prevailing before the introduction of Schematiza and Eurytoma. He concludes that the secondary succession of vegetation has now reverted to its former course.

The factors which influence the passage of flying and crawling insects through a doorway were investigated. It was considered practicable to exclude flying insects from a building or to prevent their successful escape from cages or rearing rooms by means of a downward air current if the speed of this in ft./min. multiplied by the ratio of the length through the doorway to its height is not less than 550. A combination of reflex and illuminated edges, smooth shiny walls without a pattern visible to insects, and a floor grid and underlying screening treated with a long-lasting residual insecticide, will, at all natural population densities, reduce to an insignificant level the number of insects that crawl through and survive. Artificially higher densities, as in rearing rooms, may require that the doorway be fitted with sliding doors in addition. These may be arranged to switch on the air current and lights as soon as they are opened.

A general survey of the species of Anopheles in Zaria Province, Northern Nigeria, carried out between June 1957 and December 1958, was designed primarily to ascertain the bionomics of actual and potential vectors of malaria there. The studies, which were centred upon the four principal domestic Anophelines, A. gambiae Giles, A. funestus Giles, A. nili (Theo.) and A. wellcomei Theo., were carried out by making regular floor-sheet collections and entranceand exit-trap catches in native huts, together with biting catches inside and outside huts. Data on other Anophelines and Culicines, taken at the same time, were also recorded.

The studies show that in this region, where malaria is holoendemic, all four species are endophagous to a greater or lesser extent although both A. gambiae and A. nili prefer to feed outside if a host is available. Only two species have been incriminated as vectors, A. gambiae, which had a sporozoite rate of between 4 and 7 per cent., and A. funestus with between 3 and 8 per cent., according to the season.

Collections by floor-sheets and by catches at human bait showed that A. gambiae, A funestus and A. nili were predominantly wet-season species, although in one village area studied A. funestus also occurred in fairly high numbers throughout the dry season. A. wellcomei, on the other hand, was shown to be a predominantly dry-season species.

By using traps and making collections with human bait, the entrance, exit and biting times of A. gambiae, A. funestus and A. nili were ascertained. The largest numbers of A. gambiae and A. funestus entered huts between 9.0 and 11.0 p.m. and left between 3.0 and 5.0 a.m., the maximum biting activity for A. gambiae being between midnight and 5.0 a.m., and for A. funestus between 11.0 p.m. and dawn. A. nili differed considerably from the other two species, having two peaks of maximum entry, between 9.0 and 11.0 p.m. and 1.0 and 3.0 a.m., the period of maximum exodus being between 1.0 and 5.0 a.m., with a peak of biting activity between 10.0 p.m. and 1.0 a.m. It was found that a very high proportion of the mosquitos caught leaving the huts was unfed; between 1.0 and 5.0 a.m., 64 per cent, of A. gambiae leaving, 63 per cent, of A. funestus and 30 per cent, of A. nili were unfed.

Apart from the four domestic species of Anopheles mentioned above, the only other anthropophilous species which could be described as common in the vicinity of Kaduna were A. coustani Lav., A. theileri Edw., A. flavicosta Edw. and A. rufipes (Gough). A. implexus (Theo.) is recorded from Nigeria for the first time.

The commonest species of Culicines taken at human bait during outside night collections were Mansonia africana (Theo.), M. uniformis (Theo.), M. cristata (Theo.), Aëdes lineatopennis (Ludl.) and Culex poicilipes (Theo.). Of these, M. uniformis was by far the most regular and persistent biter throughout the year. M. africana, on the other hand, was only taken in any numbers during October.

In an appendix, a list of 17 species of Anopheles (including 3 varieties), 65 of the CUlicinae and two of the TOxorhynchitinae known to occur in Zaria Province is given, with notes on their distribution and bionomics.

The name Pseudolitomastix was applied (Eady, 1960) to a genus erected for the new species P. nacoleiae, bred from Nacoleia octasema (Meyr.). This unfortunately overlooked the valid and prior use of the name Pseudolitomastix, by Risbec (1954). The two species for which Pseudolitomastix Risbec was erected are generically quite distinct from P. nacoleiae, so that a new generic name is required for this latter species. The name Pentalitomastix, indicative of the five-segmented funicle of the female antenna, is hereby proposed. I am grateful to Dr. O. Peck, of the Entomology Research Institute of the Canadian Department of Agriculture, for drawing my attention to the homonymy, and to my colleague Mr. G. J. Kerrich for suggesting the new name.

The aerial dispersal of the Pseudococcid vectors of virus diseases of cacao in Ghana and the possible influence on it of wind speed was studied at Tafo. Seasonal variations in wind speed in Ghana are slight, particularly inland in the cacao-growing areas. Observations at Tafo during 1955 showed that air movement was maximal between 12.30 and 3.0 p.m. and minimal during the hours of darkness. Daily wind speeds in the open, from April to November 1955, averaged 1·1 m.p.h., 4 ft. above ground. About 80 per cent, of wind speeds measured in one-minute periods on two days in February 1955 were within the range 1–5 m.p.h. and 11–12 per cent, between 5 and the maximum recorded speed of 7 m.p.h.

In cacao, wind speed is reduced by a factor of 10–20 times; daily speeds averaged 0·09 m.p.h., at 8 ft. above ground during April and May 1955. The highest daily averages under a closed canopy (400–650 ft./hr.) were recorded 2 ft. above ground-level; speeds fell to a minimum (25–125 ft./hr.) in the canopy at 15–25 ft., and rose again above the canopy at 30 ft. to a speed (250–350 ft./hr.) comparable with that just below the branches at 10 ft. At breaks in the canopy caused by the removal of diseased trees, daily averages were lowest (125–150 ft./hr.) 2 ft. above the ground. They rose to a peak (400–600 ft.) on a level with the lower branches of the canopy, dropped markedly (250–450 ft./hr.) at the level of the middle canopy and rose to a maximum (650–1,000 ft./hr.) at 30 ft.

All instars of Pseudococcus njalensis Laing walking on pieces of cacao wood in the laboratory withstood removal at air speeds of 10 m.p.h., but these and corresponding stages of three other species could be dislodged by gently tapping the wood. The late nymphs and adults of Ps. njalensis and Ps. gahani Green were more easily removed than their ‘crawlers’, though this difference was not found between the developmental stages in Planococcus citri (Eisso) and Ferrisiana virgata (Ckll.). Amongst the four species tested, first-instar nymphs of Pl. citri were most easily dislodged, and those of Ps. njalensis or F. virgata more so than those of Ps. gahani.

Airborne mealybugs were caught on adhesive traps, on bait twigs pinned to mature trees, and on cacao seedlings. Eight vector species became established on cacao after dispersal by air currents.

Under a closed canopy, more airborne mealybugs were caught at two and ten feet above ground than at levels in and above the canopy. At breaks in the canopy, catches averaged about 13 per cent, of those obtained under a continuous canopy and were insufficient to show changes in aerial density with height. The distribution of catches over the surface of traps would suggest that mealybugs drop from the branches, are carried by air currents when falling, and become laterally dispersed at levels a few feet above the ground.

In a clearing where cacao had been removed to simulate conditions following the routine cutting out of virus-infected trees, airborne mealybugs became established on seedlings at a distance of 45 ft. from infested cacao trees. The ratios of boxes of seedlings which became infested by aerially dispersed Ps. njalensis at increasing distances from infested standing cacao, in relation to those beneath it (unity), were: 0·86 at 10–20 ft., 0·57 at 30 ft. and 0·14 at 40–100 ft. Under conditions of high insolation, the maximum recorded distance of mealybug aerial dispersal from surrounding vegetation to cacao seedlings was 340 ft. Aerial catches on seedlings 40 to 165 ft. from cacao showed an over-all decrease with distance.

Aerial dispersal is more pronounced during dry conditions, particularly during the main dry season, December–February, and to a lesser extent during the brief dry period experienced in July or August.

The infestation rate of cacao seedlings by windborne mealybugs (predominantly Pl. citri and F. virgata) was increased by 50 per cent, when plants were protected from weather by artificial shade. These traps, insulated from ants, failed to become infested by Ps. njalensis.

Out of 64 young cacao trees, 22 per cent, became infested by airborne vectors during the five-month period May to September 1955; when, for similar trees, normal dispersal was augmented by an initial artificial colonisation with Ps. njalensis, the corresponding figure was 41 per cent. The infestation rate, after either augmented or natural dispersal, was not significantly affected by attempts to establish on the trees nests of the ant, Grematogaster striatula Emery, or by affording protection from the weather in the form of artificial shade. It was evident, however, that the presence of mealybug-attending ants is almost essential for infestations of Ps. njalensis to develop, but there must be other limiting factors, since establishment failed on 58 per cent, of trees on which coccidophilic species were present.

The part played by airborne vectors in extending infection by radial and ‘jump spread’ is discussed, together with the possible use of insecticidal measures to prevent their establishment on the trunks of healthy cacao.

In Ghana, a considerable part of the main-crop cocoa is stored for six months or more, some up to nine months, and infestation by insects increases during storage. Data are given, obtained chiefly in Ashanti on the 1957–58 main crop.

Assessment by sieving the contents of whole bags and counting the numbers of adult insects found (or, in the case of Ephestia cautella (Wlk.), the larvae and pupae also) showed that at the beginning of the main-crop season (September–November), infestation was very light (1–2 insects per bag), but was present in a high proportion of the bags (46–47 per cent, in October). The species present represent those found regularly breeding in cocoa in Ghana, namely, Ephestia cautella (Wlk.), Lasioderma serricorne (F.), Araecerus fasciculatus (Deg.), Tribolium castaneum (Hbst.), Garpophilus dimidiatus (F-), Ahasverus advena (Waltl), Cryptolestes spp., and an unidentified species of Silvanid. After six months' storage, the total count averaged over 100 per bag and comprised chiefly L. serricorne and T. castaneum; larvae of Ephestia were much rarer (less than five per bag), but adults of Ephestia were abundant in the sheds after 2–3 months' storage.

Sieving proved a useful technique for obtaining data on infestation, though careful supervision of staff is required to minimise errors. Variation from bag to bag in the same ‘lot’ of cocoa is large, and estimates based on as few as four bags are unreliable.

Assessment of insect damage on samples of cocoa beans taken fortnightly by spear-sampling showed an increase from 0·20 per cent, insect-defective beans in November to 1·0–1·9 per cent, in June. Of 21,000 beans examined in April and May, L. serricorne occurred in 0·17 per cent., and 0·6 per cent, contained living insects. Damage by insects affecting more than 2 per cent, of the beans has been very uncommon in Ghana in the past few years. The effect on quality, as assessed by commercial grading and considered financially, is at present slight, as it also is in terms of loss of weight over the whole crop (certainly less than 0·1 per cent.). E. cautella is the most important species economically, because it can infest the premises of cocoa and chocolate manufacturers in importing countries, and spread to finished confectionery. It affects a large portion of the crop in significant numbers.

The primary infestation of cocoa before it comes into licensed storage occurs in thousands of small farms and villages, and no one district or type of source is of special importance. Increase of the infestation during storage occurs chiefly as a development of this primary infestation; cross-infestation from outside sources is relatively unimportant.

The application of synergised pyrethrin in white oil within storage sheds as a fog, from which a film was deposited on the stacked bags of cocoa, was carried out in many sheds at intervals of 3–4 days, and a comparison made between sprayed and unsprayed sheds, based on data from sieving. Spraying did not substantially reduce the development of infestations of L. serricorne or other beetles, nor was E. cautella controlled to a degree comparable with experience of similar treatment against E. elutella (Hb.) in Britain; this may be due to the fact that E. cautella in Ghana develops in bags throughout the stack, and not chiefly in the peripheral bags, as does E. elutella in Britain.

Conditions affecting the film-spraying of bagged cocoa in Ghana differ greatly from those in Britain. Uncontrolled ventilation at the eaves of sheds results in a considerable loss of insecticidal fog, and uneven distribution of deposit; the undersurface of the stacks cannot be sprayed, because they rest on wooden dunnage; it is difficult to maintain an effective oil-film on the topmost horizontal surfaces of the stack, near the roof, of which the temperature frequently exceeds 100°F.; deposits probably do not retain their toxicity for more than a few days.

Variations in the materials from which cocoa-storage sheds are constructed, and in their design, cause relatively small differences in internal climatic conditions, and these have even less effect on the temperature of the cocoa in bags, which is generally in the range 82–88°F. The moisture content of the cocoa was about one per cent, less in a hotter type of shed than it was in one of a cooler type, and it is considered that the construction of sheds should aim at producing the hotter, drier conditions that are attainable, in order to keep moisture contents down and restrict the development of moulds, mould-feeding beetles and Araecerus fasciculatus.