The Tobacco Cricket, Brachytrupes membranaceus (Dru.), is a pest of young plants of tobacco, maize and other field and garden crops in Southern Rhodesia. During the period of investigations, 1949–52, it was found to be widely distributed in the Territory, the heaviest attacks occurring in areas of light sandy soil of granitic origin. All stages are nocturnal and each individual excavates and inhabits a permanent burrow in the soil, that of an adult averaging 50 to 80 cm. in depth. The burrow has an enlarged chamber in which the cricket lives and stores food.

There is one generation a year. Oviposition takes place in February and March, the average number of eggs per female, as shown by dissection, was 216. Incubation takes about 30 days and development of the young nymphs is at first rapid. From June to October, the period of dry and often scarce food, development is retarded but increases rapidly during the third and fourth (the last) nymphal stages in November. From November to May, when young and succulent food abounds, development is rapid. The adults start appearing early in December.

The eggs are laid in the burrows and the young nymphs on hatching crawl away from the parent burrow, in all directions, in search of suitable places in which to start digging. The mandibles are used to excavate the soil which is thrown outwards by the forelegs, and the mounds formed by the time the adult stage is reached may attain a height of as much as 30 cm. When an adult is evicted from its burrow, as soon as it finds a suitable place to dig, it will disappear within the space of a minute.

The burrows may be found in virgin land or in cultivated land, or along the borders of cultivated fields. The food consists of succulent or dry vegetable matter, according to the time of year, and is carried to and stored in chambers in the burrows. In the virgin veld it consists mainly of lush grasses, seedlings and root suckers of trees such as Brachystegia and Isoberlinia in the rainy season and of dried material of a similar nature during tha dry months. In cultivated land, tobacco, maize, field beans, garden vegetables, seedlings and young transplants are cut and stored.

Although the food is compacted in chambers in the burrows it has not been found in a mouldy or fermenting condition. It is thought that it is conditioned before storage, the more succulent material being allowed to wilt before being taken into the burrow and the dry material when softened by dew.

Published records and recent unpublished field observations on Trox procerus Har. in Asia and Africa are assembled. All stages of the life-cycle are figured, and brief descriptive notes on the egg, larval and pupal stages, are given. It is pointed out that the egg and larval stages have only been recorded from egg-fields of the Desert Locust, Schistocerca gregaria (Forsk.), and that the larvae feed on the eggs of the locust, sometimes causing considerable mortality. In contrast, the adults appear to survive independently of the locust, and it is suggested that the apparent concentration of adults of T. procerus in areas where Desert Locusts are ovipositing is due to their being attracted to the numerous dead locusts that are present in such areas.

Two kinds of equipment for measuring the reactions of insects to the olfactory stimuli provided by some common components of insecticidal sprays are described. One of these, consisting of a T-tube carrying a current of odorous air across the top and admitting insects up the stem, is new, can be used in the field, and gives rapid and reproducible evaluations. Tests with several species of Diptera showed that technical DDT, diesel fuel oil, and an auxiliary solvent containing methylated naphthalenes are all strongly repellent or contain strongly repellent components. Related species gave similar results with the two kinds of equipment. House-flies (Musca domestica L.) resistant to DDT were less sensitive than normal house-flies. The implications of these findings are discussed.

A method is described of estimating the total numbers and frequency distributions of adults of the Red Locust, Nomadacris septemfasciata (Serv.), in outbreak areas of hundreds of square miles, based upon counting the numbers that fly up in a two-yard strip in front of a moving vehicle. The method has proved itself valuable for indicating both immediate and future requirements for killing the locusts, but it requires refining for some research purposes.

By this method, the importance has been clearly shown of the process of congregation of scattered adult locusts in forming emigrant swarms that could start a plague. The locusts do not congregate but actually disperse just before laying eggs.

The total population in part (189 sq. miles) of the North Rukwa Outbreak Area (a self-contained area of 253 sq. miles) in Tanganyika Territory has been followed for four years. There are indications that a small migrant swarm contains 5–10 million locusts, that a total population in the whole of the North Rukwa Outbreak Area of 20 million locusts is unlikely to yield a migrant swarm, but that 50 million locusts could readily do so.

In 1953, after poor rains, no natural mortality was detected between July and October by the assessment methods described. In 1957, after good rains, natural mortality of 70–90 per cent. was revealed by the same methods, although the dry season was not fully covered by the assessments.

A colony of Eretmapodites chrysogaster Grah., sensu stricto, was maintained successfully for over a year without any reinforcement.

Identity of the mosquitos was established by starting the colony from a single batch of eggs laid by a single wild-caught female mosquito, and subsequent examination of the genitalia of the male progeny.

Mating occurred regularly after the first blood-meal of the female. In fact the females would not accept the males until they had started, or had finished, their first blood-meal.

Eggs were laid on the sterilised, water-filled bracts of banana flowers. Moist paper, moist cotton-wool or free water above paper or cotton-wool proved unsatisfactory as egg-laying media.

Larvae fed largely on the substance of the banana bracts on which the eggs had been laid. The banana bract also provided cover and so reduced the incidence of cannibalism among larvae.

Previous descriptions of the flight season of the Garden Chafer, Phyllopertha horticola (L.), are inadequate. From a study of five seasons, 1948–1952, the general or mass aspects may be outlined for the English Lake District as follows:—

The flight season starts in May or June, depending on soil temperature. When population is low the season lasts 3–4 weeks, when high it may last 5–6 weeks. Reasons for this are given.

The flight season has two overlapping phases. In Phase 1, beetles swarm close over the grass sward from which they emerged. As the season progresses, Phase 1 gradually gives way to Phase 2, in which the beetles now swarm closely on the bracken, hedges and trees surrounding the pasture. In the absence of bracken, the whole of Phase 2 takes place on hedges and/or trees. In the presence of bracken, phase-2 activity at first occurs on all three (bracken, hedges and trees) but, for some obscure reason, hedges and trees are forsaken in favour of bracken about halfway through the phase. Roughly speaking, phase-1 activity can be seen for the first two-thirds and Phase 2 for the last two-thirds of the flight season, i.e., there is about 50 per cent. overlap of the phases. The overlap is clearly due to earlier-emerging beetles entering phase-2 activity before later beetles have completed Phase 1.

Normal activity (i.e., flying closely over, or running or walking upon, the vegetation) is governed by weather. Both temperature and light are involved. Beetles are most active in warm bright conditions and, indeed, will fly and run only when the sun is shining. In warm bright-overcast conditions, they merely walk around. In cold, dull-overcast conditions (with or without wind or rain), they do not stir at all.

Activity starts any time between 8 a.m. and 2 p.m. (G.M.T.) according to weather (earliest recorded start being 7 a.m.). The first day of the flight season is usually marked by a short period of activity, 1–2 hours, in the forenoon, even when weather conditions continue favourable into late afternoon. As the days pass the period lengthens until beetles are active up to about 4 p.m. (latest recorded finish being 4.30 p.m.). The longest periods of activity, 8 a.m. to 4 p.m., occur when the sun shines all day after the first week of the season has passed. As the season draws to an end the activity period shortens until on the last day it is one hour or less. In general, rather more of the activity time occurs before noon than after.

The curve of numbers of beetles to be seen during the daily activity period is described for both phases. During the time of overlap of the phases, the curves are complicated by a certain amount of two-way traffic between grass sward and bracken (or hedges or trees). In addition, throughout Phase 2 a small proportion of individuals leaves the bracken flying high, straight and very fast, i.e., behaves abnormally; these individuals are termed “bee-liners” and will figure in a later paper.

For about the first week of the flight season all beetles disappear into the sward at the end of the day's activity. After that, a growing fraction of each day's active population spends the whole night in full view on the sward surface, bracken fringe, hedges and trees. All these “dormitories” begin to be used at the same time, i.e., about three days before Phase 2 starts. Passing the night on the sward surface ends with Phase 1. Roosting on hedges and trees ends simultaneously with activity there (first half of Phase 2). The bracken fringe alone is used right to the end of the season. Between sunset and sunrise, from 15 to 90 per cent, or more of resting beetles fall from bracken, hedges or trees to the ground. Reasons for this are given.

Beetles begin to feed when they stop spending the night below the sward surface. They feed after the day's activity is finished, and up to 8 p.m. at the latest. Their food is the herbage, bracken frondlets, leaves, blossoms, fruits, etc., on which they come to rest for the night.

In continuation of a study of the process of swarm formation in the Red Locust, Nomadacris septemfasciata (Serv.), to enable rational plans for control measures to be made, the population distribution of about 44½ million hoppers of the Red Locust was investigated in an observation area of 1,052 acres in the Iku outbreak area of the Rukwa Valley, Tanganyika Territory, from 18th January to 24th February 1957. to see if the hoppers showed a tendency to concentrate, which would have the effect of producing, immediately after the last moult, adult swarms from hoppers that were originally more scattered. From 25th January onwards, the estimated population remained constant in numbers. Densities over 50 per sq. yd. were unusual and, taking any continuous infestation over one per sq. yd. as a band, the mean density of hoppers in bands was 15 per sq. yd. There was a tendency for such bands to become larger by fusion and for the area quite free from locusts to increase slightly but the number of dispersed locusts increased. There would therefore have been no advantage in withholding insecticide control in the hope of attacking denser and more economical targets. Nevertheless, such populations have to be controlled.

In the Red Locust, under the conditions described, it seems probable that the behaviour of the young adults is most important in the formation of dense swarms, while in certain other species but not in the Red Locust, concentrated egg-laying and the behaviour of the hoppers are also important.

Investigations are required on more economical enumeration of patchy gregarious distributions.

During investigations on the phytophagous mite fauna of Mauritius, of which up to the present five species only had been recorded, 21 species were collected, of which ten species belong to the TETRANYCHIDAE. In addition, nine species, of which the majority are believed to be predacious, were observed.

An annotated list of the species, together with their food-plants, their geographical distribution and the countries from which they might have been introduced into Mauritius, is given.

Of the species of economic importance the following are worthy of mention: — In the TETRANYCHIDAE: Eutetranychus banksi (McG.) on Citrus, Oligonychus biharensis (Hirst) and Oligonychus mangiferus (Rahm. & Sapra) on various fruit trees, including avocado, mango and litchi, Oligonychus sp. (pratensis (Banks) group) on sugar-cane, Tetranychus marianae McG. and Tetranychus sp. (ludeni Zacher group) on many vegetable crops, including bean, egg-plant, tomato, etc. In the ERIOPHYIDAE: Phyllocoptruta oleivora (Ashm.) and Tegonotus sp., which are potential enemies of Citrus and tomato, respectively. In the TENUIPALPIDAE: Brevipalpus phoenicis (Geijskes) has a very wide range of food-plants and its future importance in the cultivation of tea and Citrus are mentioned, whilst Raoiella indica Hirst is an important pest of coconut plantations. In the TARSONEMIDAE: Hemitarsonemus latus (Banks) ranks as an important pest of a large number of economic plants including bean, tea and water-cress.

Only two predatory mites of great economic importance are recorded, viz.:— Typhlodromus caudatus (Berl.) and T. ovalis Evans. The bionomics of the former species are fully described and its relation to Raoiella indica on coconut is discussed. The life-cycle of T. caudatus from egg to adult varies from 4–8 days, in summer to 13–18 days in winter.

Results are given of a survey of the predatory insects associated with the phytophagous mites. The most important predators on Tetranychus spp. are the Coccinellid, Stethorus vinsoni Kapur, and the Cecidomyiid, Feltiella sp. near tetranychi Rübs. The daily mite consumption of these predators is given. Other predators of secondary importance are Oligota pallidicornis Cam. and Scolothrips sp. near indicus Priesn.

The bionomics of T. marianae on tomato and of Raoiella indica on coconut were studied. Duration of the life-cycle of the former species varies from 4–9 days in summer to 15–22 days in winter. Twenty four to 30 generations of this species are estimated to occur in a year. Raoiella has a development period of 18–26 days from egg to adult in summer and of 30–36 days in winter.

A biological study of the predatory insects, Stethorus vinsoni and Feltiella, sp., is given. Their beneficial effect is discussed. These two predators do not exert a control in time to check the build-up of the mite population. Their abundance is directly proportional to the incidence of the mite and the peak is generally only attained when the depredations of the latter have reached a disastrous level.

The various possible factors which have contributed to the sudden increase in the mite population on tomato during the last 6–8 years are discussed. Amongst the important factors are the following: —

(a) Absence of cyclonic disturbances with their heavy rains and violent gales which generally upset considerably the development of the mite population and reduce it to a minimum each year.

(b) Dressing of tomato and other vegetable crops with nitro-phosphatic fertilisers which were not applied until 10 years ago and which, it is thought, may have favoured an increase in the mite population.

(c) The abundance of natural reservoirs of wild food-plants that harbour Tetranychus spp. from which the subsequent spread results in the building up of mite colonies in cultivated areas.

(d) The ineffectiveness of Stethorus vinsoni as a predator on infested tomato plants.

A study of adult populations of Wheat Bulb Ply, Leptohylemyia coarctata (Fall.), has been carried out in the field by routine sweeping at Rothamsted. It has been observed that the males emerge slightly before the females and that the emergence period may cover at least three weeks in late June and early July. Although the number of males may exceed the number of females at first, the females predominate later in the season due to the shorter life span of the males.

The numbers of flies on the wheat have been found to fluctuate appreciably throughout the day. During the first week of the emergence period the number of flies taken increased steadily throughout the day. After the date of population peak, however, the maximum numbers occurred in the crop in the very early morning and the late evening, which suggested a daily flight dispersion followed by a general or localised return of the flies to the crop. Further study of the data showed that the daily temperature rhythm was only partly responsible for this daily flight dispersion, and that there appeared to be an active return flight to the crop in the evening. Generally the males were more active than the females and did not settle so deeply in the crop.

The temperature threshold for flight was observed to be 12 to 13°C. Winds up to 8 m.p.h. did not appear to affect flight activity, but higher winds, e.g., 15 m.p.h., markedly reduced flight, the flies remaining deep in the crop near ground level. Gale-force winds, however, were observed to produce a permanent depletion in the number of flies infesting Broadbalk field, indicating that the population was probably localised.

Although portions of the populations dispersed fairly rapidly from the emergence sites, recaptures of radioactive flies labelled with 32P indicated that the extent of the dispersions was not very great. The females dispersed more than the males, and were influenced to some extent by the occurrence of wheat in flower. Frequently the flies were found to have congregated on the lee edge of the crop, but other preferred regions have been observed which could not be attributed to the influence of the wind.

Glossina vanhoofi Henrard has been found at the I.R.S.A.C. field station at Irangi in the Belgian Congo in dense rain-forest. The fly does not normally attack man and was collected only by looking for it in its resting places. The fly rests on small saplings, head down, at an average height of about 133 cm. A three-months' survey on a fly-round transecting three valleys and two hill-tops showed that the resting density (defined as the total number of resting flies observed per 10,000 metres of path) was correlated with the profile of the transect, the hill-tops yielding more flies than the valleys. The fly-round was worked three times a day and showed the highest density during early morning in all three months; in two of them there was little difference between midday and evening catches. The resting height varied in much the same way in all the sections of the fly-round and the average height remained virtually constant during the three months of the survey. There was a slight difference between the sexes as regards average resting height, the females perching somewhat lower than the males. The sex ratio of about 1:1 was very constant throughout the day and the period of the observations.

There is some evidence which seems to indicate that this species is strongly attracted to possible host-animals by a sense of smell.

Puparia have been found in “classical” sites, but the proportion containing living pupae was small. The duration of the pupal period is about 35 days under laboratory conditions at 26°C. and 90 per cent. relative humidity.

G. vanhoofi in captivity feeds readily on the porcupine, Atherurus africanus.

During the breeding seasons 1952 to 1955, samples of females of the Red Locust, Nomadacris septemfasciata (Serv.), were caught in the Rukwa Valley, Tanganyika Territory, and the ovaries classified according to egg length. Ovaries designated Class V were those regarded as containing eggs ready for oviposition. Population estimates and rainfall records were also taken.

The data suggested that the eggs in the ovaries develop very rapidly, becoming ready to be laid about seven days after the start of development.

The increase in weight of an individual female during ovary development is related to absolute size, and it was found that, during ovary development, the weight increases by approximately one-half.

The most important factor initiating ovarian development seemed to be a drop in daily maximum temperature to below 36°C. The factors of high humidity and rainfall were of less importance but their influence is not fully understood. The possibility of change in length of day being an important factor in the initiation of ovarian development is discussed.

The limiting factors for the number of egg-pods laid in a season are considered to be the dates of the drop in daily maximum temperature and of the drop or decline in parental population. The date of drop or decline in parental population seemed to be connected with the date of drop in daily maximum temperature.

The percentage of females laying more than one egg-pod was calculated, and the resuJts were found to vary considerably from year to year, although the average of 1 pods per female agreed with past estimates. Also, the average period of 16 days between successive egg-layings agreed with past work.

Drought, during the laying period, seemed to cause the females to retain their ripe eggs.

The evidence tended to confirm the possibility of a Red Locust population being able to increase 100-fold in a single year.

Cage experiments confirmed that, in the absence of natural enemies, populations of Aphis fabae Scop., attended on bean plants (Vicia faba) by the ant, Lasius niger (L.), multiply more rapidly than otherwise similar but ant-free populations. The average difference in numbers recorded, was about one-third, the maximum being 70 per cent. No doubling or trebling of aphid numbers as claimed by an earlier worker was ever recorded.

When the Aphids are attended by ants, their excretion behaviour alters and the normal dispersal of the apterae from the young apical growth of bean plants is considerably delayed.

No significant differences were found between the numbers of nymphs produced by individual Aphids from ant-visited and ant-free plants, respectively, living on leaves of the same age; but the numbers were significantly affected by the age of the leaf or part of the plant on which the Aphids had developed or were then feeding.

It is suggested that ant-attended aphid populations multiply more rapidly because most of the Aphids feed for a much longer time on young plant tissue where, presumably, their food supply is more nutritious.

A campaign in New South Wales against flying swarms of the Australian Plague Locust, Chortoicetes terminifera (Wlk.), using Tiger Moth light aircraft, is described. Large rolling swarms of this locust, travelling across the open plains of western New South Wales, are directed in their general south-easterly move ment by natural barriers in such a way that regular migration routes are followed and predictions of swarm movements can be made. Of particular importance are the timber fringes of creeks and rivers which tend to baulk swarms temporarily and it was in such situations that locust swarms proved most susceptible to treatment by a manoeuvrable aircraft.

A technique of drift spraying was developed wherein aircraft were able to operate without ground direction against travelling swarms by spraying at right- angles to the wind direction and by using the main body of the flying locusts for demarcation. Large numbers of short spraying runs were made in rapid succes sion over the most dense portions of the swarm being treated until control was achieved. It was found that the spacing of the runs was not particularly important as long as each was applied to the densest portions of the swarm. The light and relatively slow-moving Tiger Moths proved highly suitable for this work and they were also useful for scouting and keeping track of swarms. Observations made after spraying indicated that the coherent flying swarms treated in the open plains were almost completely eliminated. If, however, swarms were able to reach closely settled or heavily timbered areas they tended to split and disperse and were difficult to treat. The insecticide used was an emulsifiable concentrate containing 7 per cent. γ BHC, which was diluted with diesel oil, and one gallon of liquid containing 5·6 oz. of γ BHC was designed to cover one acre. However, as the campaign progressed, effective use of the spray was probably obtained at less than one gallon per acre, and a total area of 30–37 square miles of dense swarms was treated at a cost of approximately £13,700. There is scope for an improvement in the efficiency of the spraying technique as well as a reduction in some of the major costs in any future campaign.

It was considered that the campaign carried out in the Ivanhoe—Hillston district completely controlled all swarms in the area and protected from invasion valuable farming lands to the south-east. Further spraying operations, in the Jerilderie district, 150 miles away, were less successful, although many mor locusts were killed. The spraying party arrived in the district too late to plah a strategic campaign and some swarms escaped treatment.

It is concluded that the method of aerial drift spraying is capable of wider application in Australia, particularly in areas where problems of distance, popula tion and lack of facilities militate against successful control by ground machinery.

Dr. W. Thalenhorst, of the Niedersächsische Forstliche Versuchsanstalt, Göttingen, Germany, sent me some time ago specimens of what he took to be an abnormally dark form of Pachynematus montanus (Zaddach). These had been reared in 1952 from larvae, not at the time distinguished from the larvae of that species and found feeding with them on spruce (Picea abies) in the Harz Mountains in the previous year. Unfortunately the female was smashed in the post and I did not like to describe a new species that would have to be based on only a few males, though I was not able to identify them with any known species. In 1955, however, he reared another female of the same species and from this he tried to obtain eggs but without any success. This female I now have before me. No further specimens have been found.

An experimental control project against Simulium damnosum Theo. in Northern Nigeria using DDT in diesel oil as a larvicide is described. A description, together with a map, is given of the project area, some 1,200 sq. miles in extent, with a population of 32,000 persons, and it is emphasised that this was selected to provide information on the feasibility of control in an area not isolated from other fly foci.

Methods used in applying the larvicide to the rivers, which had rocky beds with rapids and falls, and assessing the effects of treatment are discussed. Applications were made at prearranged treatment points, weekly, for a period of 12 weeks in the dry season of two successive years, 1956 and 1957, and the post-treatment density of adult fly was assessed and compared in a Table and in histograms, with that existing in the year prior to treatment.

In the first year, a mean dosage per application of 1·4 p.p.m. of p,p′DDT applied over 30 minutes (S.D. = 0·5 p.p.m.) at each of four treatment points resulted in the clearance of all breeding for a distance of 15–23 miles below these points, and a reduction of adult fly density in the area of 96 per cent. as compared with pre-treatment density. But reinfestation of cleared breeding grounds was found to be rapid, these becoming repopulated with immature stages within six months in spite of a mean wet-season fly density of only 0·35 flies per boy-hour.

In the second year, with application of 1 p.p.m. of p,p′DDT for 30 minutes, results were less successful than in the first, although three additional treatment points were used with the aim of controlling a greater distance than in the first year, adult density being only reduced by an estimated 82 per cent. of that before treatment in the first year. The reasons for this are discussed, and the lower degree of control is attributed to the partial failure of the insecticide to reach the breeding sites owing to the exceptionally low state of the rivers.

The fly-rounds used in assessing adult density and the pre-treatment breeding sites and the positions of the treatment points are shown on maps. Detailed data from the fly-rounds are given in an Appendix.

With the exception of slight mortality among fish, thought to be due to accidental impediments to tbfi free downstream flow of the larvicide, no toxic effects on other organisms were observed.

The cost of treatment is given, and the results are discussed briefly in relation to the existing knowledge on control of S. damnosum. It is suggested that control by the application of larvicides in non-isolated foci is a practicable possibility, but that annual treatment might be necessary because of rapid re-establishment in cleared areas. The effect of control, of the degree obtained in this work, on the incidence of human onchocerciasis cannot yet be determined.

During the summer of 1956, a quantitative study of a field population of Wheat Bulb Fly, Leptohylemyia coarctata (Fall.), was carried out at Rothamsted.

The work consisted principally of a study of the development and decline of a population of adult flies. This was supplemented by observations on the populations of the immature stages.

Emergence was investigated by the use of a cage of fine terylene netting, 24 ft. long, 12 ft. wide, and 6 ft. high. This was erected in an infested wheat field shortly before flies were expected to appear, and was searched twice daily, at 10 a.m. and shortly before sunset.

A total of 293 male flies was caught during the 26 days from 24th June to 19th July. Of these, 186 appeared between 1st and 6th July. The highest day's catch was 40 flies on 6th July, by which date (inclusive) about 84 per cent, of the final total had emerged.

A total of 258 female flies was caught during the 35 days from 21st June to 25th July. Of these, about half emerged between 6th and 10th July. The highest day's catch was 37 flies on 8th, by which date (inclusive) 66 per cent, of the final total had emerged.

Population decrease was investigated by the method of marking and recapture. The newly emerged flies caught in the cage were marked with Artist's oil colours and released in the cage. The colour of the mark was changed daily so that the age of marked flies could be ascertained. A search was made for marked flies every three days and their numbers, marks and sex were recorded. From the recapture figures, estimates of the numbers of flies surviving at different times after marking were obtained. Mortality during the first day was very high, but after this numbers decreased at a steady rate. This initial high mortality was believed to be due to marking. The length of life of marked flies which survived this immediate effect was, however, not impaired, therefore the rate of population decrease was estimated from the recapture figures alone, that is, without reference to the numbers originally marked. The half-life of male and female populations was estimated as 7·3 and 11·1 days, respectively.

Application of the estimated rate of population decrease to the observed emergence figures enabled a general picture of the size and structure of the population to be obtained. The predominance of males over females during the early part of the season and the later predominance of females over males were explained.

Observations on the populations of the various stages showed that the mortality between pupation and maturation of adults was high, and that the egg populations inside the cage during the autumn of 1956 was only about 1/7th of that of the previous year. This reduction was not observed outside the cage and may have been due to the survival rate and fecundity of the flies being affected by the experiment. Further work will be necessary before this can be elucidated.

The departure of the newly emerged adults of the salt-marsh mosquito, Aëdes taeniorhynchus (Wied.) from the breeding sites was studied in Florida. The departures, which are presumed to be the initial stage of the migratory flight, have been observed to take place only in the dark period of the 24 hours, and, the vast majority of individuals being ready to take off at dusk, spontaneous mass departure took place at that time.

Samples of the resting population of mosquitos were taken at intervals before and after the mass departure, and from the departing migrants. The average age of the males in each sample was estimated by the proportion in different stages of hypopygial rotation, from which it is concluded that migratory activity begins when the insects are six or more hours old. The end of the migratory period is not yet known. In all of 21 female departing migrants examined, the ovaries were immature and the spermathecae empty.

The efiect of the previous crop on the subsequent infestation of winter wheat by the Wheat Bulb Fly, Leptohylemyia coarctata (Fall.), mainly on the clay-loam of Rothamsted Experimental Farm has been investigated for the years 1954 to 1956. By far the highest larval populations followed fallows, the next highest potatoes, which are a low, open cover crop, whilst much lower populations followed the tall crops of beans and wheat, and very small infestations followed the grass mat covers. This confirms the results of previous workers under different agricultural conditions.

Previous applications of dung and straw influenced the infestation following fallow in dry summers but not in a wet year. The effects of other manurial treatments were inconclusive.

An experiment designed to determine the effect of the crop density on oviposition was inconclusive due to the failure to establish an infestation.

Wheat was shown to influence the flight path and to interfere with local egg-laying behaviour, producing a horizontal effect of up to twice its own height.

Apart from the foregoing factors, at Rothamsted, where the wheat fields are well separated, the differences in the level of infestation that were observed could be explained by close proximity or otherwise to a centre of heavy fly infestation.

Reasons are put forward for suggesting that the effect of the previous crop on oviposition may be interpreted in terms of the opportunity existing for contact with the soil rather than as a preference on the part of the fly.

In view of the discovery of several new species and subspecies of Amblypelta (Hemiptera, COREIDAE), some of which are either known to be or else threaten to become of economic importance, a revision of the whole genus has become necessary.

The genus Amblypelta, as defined in the present paper and as known so far, is confined to Australia north of Sydney, the islands between Australia and New Guinea, New Guinea and, to the west, the Kai Is., Timor and Java; the Bismarck Archipelago, the Solomon Islands including Bougainville, Eennell and Bellona; the New Hebrides and New Caledonia. In the present revision of the genus, twelve species and five subspecies are included, of which all but seven species are new to science. Apart from the descriptions of the new species and subspecies, taxonomic notes on the previously known species are given as well as a key to all the species.

The known distribution of each species is stated, and consideration given to the zoogeography of the genus as a whole, and of certain species-groups which can be recognised within the genus.

Twelve species of termites occurring in the British Solomon Islands Protectorate are listed, and an illustrated key to their soldier castes is provided. An analysis of the 114 records has been made in order to indicate, in a general way, the relative abundance of the various species. Notes are given on distribution within the islands, on habitats and nests, and on the occurrence of termites on coconut palms and in buildings. Damage to buildings is discussed and suggestions made for control measures applicable to local conditions.