The biology of Melittomma insulare Fairm. (Coleoptera, Lymexylonidae), a wood-boring pest of coconuts in the Seychelles and northern Madagascar, has been studied from all angles, with the object of discovering practical means of control.

The stages in the life-history are figured and briefly discussed, supplementary details being given to the earlier accounts of Vesey-FitzGerald and others.

The feeding habits of the larvae ae described in some detail, with particular reference to their association with micro-organisms which were almost invariably detected in infested wood surrounding the tunnels. The larva is shown to feed on fluid extracted from the parenchymatous tissues and not on solid wood.

The course of the development of an infestation in an individual trunk is described in detail, culminating in the fall of the palm. Variations in this process are correlated with the age of the palm and other factors. The spread of infestations through plantations and islands is also discussed. A detailed account is given of the symptoms shown by attacked palms, together with a critical revision of certain changes which have in the past been attributed to Melittomma attack.

An account is given of the occurrence of M. insulare in certain palms other than coconuts, in which it has been found.

A series of 32 cloths, including 8 flannels dyed to Munsell standard colours, seven coloured satins and 17 service fabrics, were tested for their attractiveness to northern Canadian Aëdes mosquitos, and were assessed for their reflectivity of visible and ultra-violet light.

Their attractiveness was found to vary inversely with their reflectivity or brightness, although the different textures represented in the series tended to obscure the generalised relationship.

When comparisons were made within a series of identical texture, the correlation was good in the range between 475 and 625 mμ wave-length, but was not present in the red or infra-red, nor in the violet and ultra-violet ranges of wavelength.

Green tended to be less attractive, and white more attractive, than was expected from this relationship. For colour values of similar lightness or darkness, the order of attractiveness to Aëdes punctor and associated species was black, red, blue, brown, green, white, yellow.

An investigation has been made of the deposition in open country of a coarse aerosol released from an aircraft flying at approximately 30 feet. A constant drop spectrum was used, with a mass median diameter of approximately 80 microns.

It is shown that the rate at which the aerosol deposits is greatly affected by the degree of atmospheric turbulence, a practical measure of which is described. The rate of deposition is approximately inversely proportional to the time interval elapsing between production of the aerosol and deposition, except for deposits laid down near the line of emission.

Droplets less than 10 microns in diameter are not deposited appreciably within any useful time interval, and it would be most uneconomical to attempt to produce deposits upon the ground by using drop spectra consisting mainly of such small droplets.

The solution used was a 10 per cent. w/v solution of technical DDT in a solvent consisting of 1 part of Shell Power Kerosene to 1 part of Shell Diesoline. Evaporation of the volatile constituents takes place very rapidly during the first few seconds after production of the droplets, and deposits formed immediately beneath the aircraft track have lost 50 per cent of their initial volume. Solvent losses increase significantly with time, but so slowly that for most practical purposes the solution strength of the ground deposit is very nearly constant for all important sampling points.

In British East Africa Oecophylla longinoda (Latr.) var. textor Santschi is locally common in the costal region. Inland it is absent from higher altitudes and from areas where there is a pronounced dry season.

In Zanzibar Island, O. longinoda at least 89 species of trees and shurbs; the largest populations occur on the clove (Jambosa caryophyllus), Citrus spp., Bridelia micrantha and Canthium zanzibaricum.

The nesting habits and colony composition of O. longinoda are such that one colony may spread over a number of adjacent trees; it contains only one gravid queen.

Winged virgin sexual forms are released at the beginning of the wet seasons and new colonies are initiated by a single queen, who uses her food reserves to bring the first batch of brood to maturity.

In Zanzibzr, O. longinoda tends a wide range of Homoptera that produce honey-dew, but apparently “ prefers ” certain Coccids, notably Saissetia spp.

The degree of attention afforded by an ant species determines the species of Homoptera which it is able to attend.

The insect species preyed upon by O. longinoda include the honey bee, Apis mellifera, and the driver ant, Dorylus nigricans, of which large numbers may be destroyed.

O. longinoda is of undoubted value for controlling certain coconut pests, notably Theraptus sp. (Coreidae), and its efficiency in coconut plantations could probably be much enhanced.

A close association exists between the ant Oecophylla longinoda (Latr.) and the scale insect Saissetia zanzibarensis Williams.

Several factors cause the scale to be rare in the absence of the ant. Contamination by honey-dew and sooty moulds prevents increase of the S. zanzibarensis population above a relatively low level, while, in addition, insect parasites and probably predators virtually exterminate the scale. A fungal parasite may destroy honey-dew-contaminated S. zanzibarensis under humid conditions.

The attendant O. longinoda prevents contamination of the scale by honeydew. It gives absolute protection from Coccinellid predators and, although parasitism by Coccophagus spp. and predation by Eublemma spp. is not prevented, the mortality of S. zanzibarensis which they cause is thought to be of little significance.

O. longinoda workers benefit S. zanzibarensis in other ways. They remove débris from the scale clusters; they transport the nymphs and establish them at suitable feeding sites. The effect of these benefits is most marked when the ant population is high relative to that of the scale.

The silken shelters built by O. longinoda over S. zanzibarensis clusters protect the ant during adverse weather and may only incidentally benefit the scales.

S. zanzibarensis and other honey-dew-producing Homoptera are the major food source for O. longinoda. Other sources are sufficient to maintain only relatively low populations of the ant.

In an earlier paper (Woodroffe, 1953) full details were given of the results of a survey, the purpose of which was to study the insects and mites which inhabit certain birds' nests in Britain. Since then further work has been carried out on certain species, and, as a result, one or two statements made in the earlier publication now require correction or modification.

Three volatile solid insecticides (DFDT, γ BHC and aldrin) were tested against two species (Oryzaephilus surinamensis and Tribolium castaneum) at two temperatures (30°C. and 11°C.) in simple laboratory fumigation tests. DFDT killed both species faster at 11°C. than 30°C.; γ BHC killed O. surinamensis faster at 11°C. than 30°C., but T. castaneum faster at 30°C. than 11°C.; aldrin killed both species faster at 30°C. than 11°C.

The Makueni fly belt is chiefly infested with Glossina pallidipes Aust. and lies between 4,000 and 2,600 feet above sea level. The climate is hot and dry with two wet and two dry seasons, annually. The dry period during January and February is hot with bright sunlight while during the June to September dry period the temperatures are lower and with dull cloudy periods. The vegetation is Acacia-Commiphora dry grassland association with the Commiphora especially predominating at the lower altitudes in the east. The non-riverine deciduous bushland and thickets show two leafless periods in the year, corresponding with the dry seasons. A system of regular fly patrols has been in operation in this area since January 1949. The total flies caught show a regular seasonal increase during February-March each year and less regular and smaller increases during June-July and October-November. The same seasonal increases in the fly-populations are shown by the average number of old male flies caught per patrol.

Flies caught during 1951 were fewer than in 1950 in the settled areas and at Thwake camp, but at Kikumeni in the dense Commiphora-Acacia thickets the catches in 1951 were at least as numerous as in 1950.

A breeding-place of Glossina pallidipes Aust., in evergreen shrubs of Craibia sp. on the Kiangine river, Makueni, Kenya, was studied by means of hand-searching for puparia in conjunction with records of rainfall and temperature. A fly-patrol was also operated. The country away from the river was thorn-scrub, with Acacia and Commiphora trees predominating.

Three sites, each of 200 sq. yards, divided into 50 squares, were marked out, and two of them were searched monthly and the third quarterly.

Site “ A ” was searched nearly every month between April 1949 and June 1952, and full puparia were nearly always found. Numbers varied very widely; and it was noted that few could be found in rainy seasons during the months of November to January and April to May, though 1950 was an abnormal year. A consistently large increase in numbers was observed in February, the hottest and driest month, and a rather less marked increase towards the end of the long dry season in September and October. Large numbers of full puparia also tended to coincide with times of low average minimum temperature.

The numbers of flies caught on the fly-patrol did not appear to bear any close relation to the numbers of full puparia discovered.

Testing for efficiency of searching by means of marked puparia showed that about 40 per cent. of puparia were recovered, so that results must be taken as relative.

1. Disappearance of Simulium neavei after DDT-treatment of rivers in Kodera, Kenya, is still manifest after seven years, and the species seems to have been eradicated from the district.

2. Skin microfilaria rates have dropped from something under 7 per cent. to nil in the under 4 years age group; from 37 to 5 per cent. or less in the 4–8 years; and from 70 per cent. to 50 per cent. in adults.

3. The length of life of the adult worm can be at least seven years.

Measurements of the Horizontal Equivalent Area (HEA) of flying locusts and the toxic efficiency of Mk. IV DNC solution, as a function of spray droplet size, have been made using Schistocerca gregaria (Forsk.) flying “ naturally ” in a wind tunnel. The experimental conditions, including visual stimuli, for “ natural ” flight were determined by means of a flight balance. The observed values of locust airspeed varied from 2·5 to 6·5 metres per second with a mean value of about 3·2 metres per second.

It is felt that the 24-hour catch presents a valuable method of studying the biting-behaviour of mosquitos in the field, and further that conclusions concerning seasonal abundance, vertical distribution and times of biting-activity must be based on catches of this type if serious errors of interpretation are to be avoided.

The time-divisions of the catch should not exceed one hour, and even shorter intervals may be desirable. Timing should be related to the actual times of sunrise and sunset. Where possible, a series of consecutive catches should be carried out, and a shift system must be carefully considered in relation to the particular series projected.

Usually the results from different levels above ground should be treated separately in working out biting-cycles, and this may also apply to results from different seasons or different localities. Further, while in some instances the summation of long series of catches seems permissible, in others it is not, and in these detailed analysis of the figures may be necessary before consistent behaviour-patterns become apparent.

In cases where a measure of the central tendency is to be employed, the geometric mean as modified by C. B. Williams appears to be the most suitable.