The coconut scale, Aspidiotus destructor Sign., was recorded for the first time from the island of Principe in November 1954, but it is practically certain that it was actually introduced accidentally in 1952.

Prior to 1954, normal fluctuations had occurred in the total annual production of copra of the island, probably correlated mainly with rainfall conditions. However, in 1954 there was a steady and very serious decline in production due to damage caused by Aspidiotus. Several species of endemic natural enemies were found attacking the scale, but the control exerted by them was quite inadequate to prevent astronomical increase of Aspidiotus.

The following species taken in London on imported bananas has been submitted recently for identification and has proved to be a new species of Dysmicoccus Ferris.

On checking some unidentified material in the British Museum (Natural History), the same species has come to light on imported bananas collected in London earlier. As some of these were labelled “from Canary bananas”, a check was made of all the material in the British Museum (N. H.) of mealybugs from the Canary Islands and it is evident that the species is widespread there but has been misidentified as Pseudococcus comstocki (Kuw.). Furthermore, it has been misidentified as this species on other occasions on imported bananas in England and Egypt and although some of the preparations concerned are labelled as from Canary bananas, it is not possible to determine whether all have come from this source.

In the course of an experiment in malaria control in an inland region of Kenya and Tanganyika, by the use of house spraying with dieldrin, routine catches were maintained of mosquitos resting in artificial outdoor shelters. During the 18 months of the pre-spraying period, catches in the South Pare district of Tanganyika mainly consisted of the principal vectors, Anopheles gambiae Giles and A. funestus Giles, together with small numbers of A. rivulorum Leeson.

During the three years following the spraying, A. funestus disappeared almost completely from the catches, while A. rivulorum showed an increase of about seven times above its former level.

A stretch of river with its flood plain, a habitat of Glossina morsitans submorsitans Newst. and G. tachinoides Westw. in the Sudan savannah zone of Northern Nigeria, which is an important cattle-raising area, is described.

An effective technique for the eradication of both species of tsetse has been developed, using a single application of a 5 per cent, aqueous suspension of DDT by knapsack sprayer in the places of dry-season concentration of the fly, and on the principal midday resting sites, viz., between ground-level and a height of 5 ft. on tree trunks greater than 6 in. in diameter.

An account is given of work carried out during the third year to eradicate Glossina morsitans submorsitans Newst. and G. tachinoides Westw. from an elongated fly-belt situated in the Sudan savannah vegetational zone of Northern Nigeria, which is an important cattle-raising area.

The total area of the fly-belt, which is isolated as far as G. morsitans is concerned, measures about 120 miles in length and extends, in places, up to almost 10 miles in width. The country involved forms the flood plains and adjacent uplands of the Komadugu Gana river.

Sixty-nine sq. miles were sprayed with DDT during the 1957–58 dry season(between end of January and end of April 1958), and 18 months after completion of work no tsetse has been found in the treated area. These 69 sq. miles formed the dry-season habitat of the fly on this section of the river, and the cost of insecticide and labour involved amounted to approximately £86 per sq. mile. As the zone infested in the wet season greatly exceeded this dry-season concentration area, reclamation costs per sq. mile, when applied to the amount of grazing land made safe for cattle, amounted to much less than the figure quoted.

Successful continuation of this project is ascribed to three salient features of the technique employed: —

(a) A single application only, of a 3·75 per cent, aqueous suspension of DDT from a wettable powder, is sufficient for complete eradication.

(b) A high degree of discriminative or selective spraying is possible: for the elimination of G. morsitans, spraying can be mostly confined to larger tree trunks, in shade, up to a hei ght of about 5 ft.

(c) Artificial or natural barriers to isolate each season's work, and so prevent re-invasion, are not necessary where the fly-belt is of a comparatively narrow and elongated nature. Spraying the re-infested area during the following season is more economical.

Because the lettuce root aphid, Pemphigus bursarius (L.), is gall-inhabiting on poplar in spring and lives underground on lettuce during summer, its naturalenemy complex differs from that associated with foliage-feeding aphids. It suffers little from parasites and only fundatrices have been found attacked. The only species reared from parasitised individuals was the Pteromalid, Pachyneuron sp., which is assumed to be hyperparasitic. When the galls open at migration time, Syrphid larvae and Anthocorids can enter and attack the contents. Anthocoris nemorum (L.) and A. nemoralis (F.) are the main predators at this stage. Before they begin feeding, they kill all the aphids within a gall extremely quickly and efficiently. It is suggested that the aphids may be killed by a secretion from the stink glands of the Anthocorids working as a fumigant within the confines of the gall. Several Staphylinids and Carabids are frequently associated with subterranean colonies of P. bursarius and are almost certainly predators. The larvae of the Chloropids, Thaumatomyia glabra (Mg.) and T. notata (Mg.), are common predators of P. bursarius in the soil. The sexuparae, both before they leave the lettuce plants and on their arrival on poplar, are attacked by three species of Coccinellid.

Anthores leuconotus Pasc., indigenous in East, South-East, South and South-West Africa, is an important ring-barking and wood-boring pest of the main stem and root of arabica coffee. The duration of the life-cycle ranges from 12 to 25 months, with most individuals requiring 16–20 months. Thus beetles developed from eggs laid during the long rains (April–June) of one year will mostly emerge during the short rains (Nov.–Jan.) of the following year, with some appearing during the rains preceding and others during the ensuing rains.

If the mean life-cycle is taken to be 18 months, the approximate duration of each stage is as follows:— eggs require 21–23 days for eclosion; larvae in the first five instars ring-bark the tree for 4 months, and those in the final two instars bore into the wood for 12 months; the pupal period lasts about 4½ weeks, and the beetle remains in the tree for a further two weeks before emerging.

A survey of the parasites of aphids that occur on lettuce was made in two localities, in the Midlands and northern England, respectively, and a list of these parasites compiled.

A study was made of the Braconid parasite, Monoctonus paludum Marshall, in relation to its host, Nasonovia ribis-nigri (Mosley), on lettuce. Techniques for observing the parasite's behaviour towards various species of aphids are described and methods of conducting experiments on breeding success in species other than N. ribis-nigri are given.

Trials in East Africa (chiefly in Tanganyika) are described to show that the time taken for first-instar larvae of Busseola fusca (Fuller) to reach the more closely packed young leaves in the central axis of the maize plant is about ten days, and that insecticides are effective up to the end of this period. Results show that the best time for control is as near hatching of the eggs as possible, but with more persistent insecticides such as endrin, economic control is possible up to five days later.

A method of testing insecticides as residual deposits, by allowing first-instar larvae to walk on them, is given, and it was found that isodrin, DDT, endrin and γ BHC are highly effective as residual films. The MLD's for one minute in contact are 1·8 mg./sq.m. for isodrin, 4–7 mg. for DDT, 8·3 mg. for endrin and about 20 mg. for γ BHC. Derris emulsion has no residual effect after drying and is only effective in the liquid state after contact with larvae for several minutes.

In a series of field trials over three years, increases in yield of up to 2·6 times the control yields were shown to be possible, and averaged twice the control yield. Endrin was the most consistently effective insecticide, applied as a 2 per cent, dust and as a 0·03–0·04 per cent, emulsion spray. Two applications of endrin at an interval of two weeks were effective, but, for DDT, three applications of an emulsion containing 0·05 per cent. DDT at nine-day intervals were more effective than two while one may be of little value. An application of 2·5 oz./acre of DDT can be effective for 7–10 days, while 1·5 oz./acre of endrin can be effective for 14 days. In general, however, rates per application should not fall below 3·5 oz. DDT or 2 oz. endrin per acre, overall.

The superiority of a 5 per cent. DDT dust over one of 2·5 or 1 per cent, makes it more economical in a high-yielding, heavily attacked area. Diazinon and malathion in emulsion sprays are less persistent than DDT.

Trials with 4 per cent, malathion dust and 0·15 per cent, malathion emulsion spray were inconclusive on account of heavy damage by Spodoptera sp.

Attempts to increase the persistence of malathion and γ BHC by the addition of chlorinated terphenyl resins were also inconclusive, but DDT in emulsion, with resin, appeared more effective than DDT alone.

The economics of successful control measures at the prices obtaining at the time are discussed under the various trials.

Chemical estimations of the deposits of insecticides retained on the plants showed that between 1·4 and 3·7 per cent, of the quantity applied per unit area can be recovered in this way.

In a discussion of the significance of the results in control, it is suggested that spraying the stems only is half as effective as spraying the whole plant.

The conditions and results relating to two aerial spray trials are given.

The place of control of Busseola in agricultural practice is described, and some of the present results are quoted to show that yields can be increased by good farming practice as much as by control of Busseola.

A general description of mass behaviour of the garden chafer, Phyllopertha horticola (L.), in the flight season was given earlier in this series (Milne, 1958). The outstanding features are two roughly equal and half-overlapping phases of activity: Phase 1, the swarming of the beetles over the grass sward, followed by Phase 2, the swarming on the surrounding bracken, hedges and trees. Investigation of sex proportions and mass reproductive state in the two phases permitted some deductions as to individual male and female behaviour (Milne, 1959a, b). In the present paper, further details of behaviour are given, and study of the flight season is concluded with a brief general discussion.

Methods that have been used to confine insects to the test surface during exposure to insecticidal deposits are briefly reviewed, and their limitations are discussed. Some unsuccessful preliminary experiments on alternative methods are mentioned.

A temperature-preference method of inducing house-flies, Musca domestica L., to remain on a treated surface is described, and the apparatus is illustrated. The treated surface forms the floor of a chamber; it is maintained at a higher temperature than the walls and lid. The flies are introduced by means of a plunger device.

The true identity of the specimens from Africa recorded as Coccus africanus (Newst.) and C. viridis (Green) is reviewed in detail. Four other species of green scales attacking coffee in Africa south of the Sahara are described as new. These are: C. alpinus—for more than 40 years confused with africanus—C. celatus, C. consimilis and C. viridulus.

In pot experiments, the ability of newly hatched larvae of the wheat bulb fly, Leptohylemyia coarctata (Fall.), to move through nine in. of soil and infest wheat plants was tested for three different types of soil: sandy loam, clay loam and peaty loam and also clay loams at different levels of acidity. Each soil was inoculated with 200 freshly hatched larvae and the effects associated with soil were measured by the number of larvae subsequently recovered from infested plants.

The larval recoveries for different soil types were: sandy loam 28 per cent., clay loam 19 per cent, and peaty loam 2 per cent. The larvae could infest plants grown in clay loam within a pH range of 4·9 to 7·8 and the results suggested a possible optimum at pH 6·2. The relative failure to infest in peaty loam appeared to be due to this soil impeding larval movement rather than interfering with the mechanism of host-plant location.