The considerations affecting the design of a spray testing chamber are discussed, and a description is given of equipment specially intended for use with mosquitos as the test insect. In essentials the apparatus consisted of a metal chamber surrounded by a “Masonite” cabinet with an air space in between. In this space air, at a constant temperature, was circulated around the outer face of the inner metal chamber and so controlled the temperature of the spray chamber. The cabinet was fitted with lights; wet and dry bulb thermometers to check the humidity, which was controlled by blowing in steam; a fan to circulate the air; a spray gun; four ports through which the insect exposure cages were introduced into the chamber; a shelf to carry these cages; and an exhaust and an intake port to allow the mist to be removed at the conclusion of the test. Finally, there was a door on one side allowing access to the interior. It will be noted that the insects were introduced into the spray chamber in cages and not allowed to fly freely. This greatly facilitated the collection of insects at the conclusion of the test. Under the routine testing conditions adopted the insects were not introduced until four minutes had elapsed after spraying. They were then exposed for ten minutes and afterwards transferred to clean recovery cages.

The practice of caging the insects and placing them on a shelf opposite a window within the spray chamber permitted the insects to be closely observed and the rate of paralysis (knockdown) to be followed. In general, however, the rate of knockdown, when it occurs, was not so sensitive a method of comparing insecticidal efficiency as the twenty-four hour percentage kill.

In practice the apparatus was found to give satisfactory results. It permitted accurate comparisons to be made between different strengths of a given insecticidal solution, but care had to be taken to ensure that the non-volatile content of the sprays being compared for toxicity was essentially similar. When this was not so, other things being equal, the spray which produced the less volatile mist (i.e., the larger particle size) gave the higher kill. When working on activators, which are usually non-volatile materials, it was therefore necessary to add an equivalent quantity of, say, high vacuum pump oil to the control spray before making the comparison. Examples of the application of the method of pyrethrum, DDT, and activated sprays are given.

The set of standard conditions chosen for the routine testing procedure were sometimes rather arbitrarily fixed but usually with due reference to the requirements of the insects or to practical testing conditions. Since two of the conditions chosen, i.e., the use of caged insects and exposure after an interval of four minutes after spraying, were rather artificial from the point of view of practice and might conceivably have led to conclusions inapplicable to practical conditions, the effect of modifying certain of the standard procedures was investigated. The results of these experiments are fully described in the text. It was concluded that temperature influenced the level of kill recorded considerably, the kill being lower when the temperature of the spray chamber was dropped 10°C. Humidity seemed to have little or no effect within the range 40–80 per cent. R.H. Increasing the interval at which the insects were inserted after spraying from 0 to 10 minutes affected the level of kill without influencing the comparative efficiency of the insecticides examined. On the other hand caging the insects had a pronounced effect on the comparative efficiency of pyrethrin sprays with and without added non-volatile material. This was thought to be due to the delay in access of the spray droplets to the caged insects which permitted a sufficient interval for the differences in particle size to become established. With caged insects inserted before spraying, therefore, the biological efficiency of a pyrethrin spray without, added non-volatile material is much lower than when some such substance as high vacuum pump oil is added, whereas this difference is much smaller when the insects are exposed free in the cabinet (Table X).

1. 97 per cent, of the ♀ A. gambiae, A. funestus and A. melas remained until after dawn in the huts in which they had fed.

2. ♀ A. melas seldom left suitable huts in the daytime.

3. There was a high positive correlation between the time at which ♀ A. funestus entered huts and the time of moonlight. The maximum correlation coefficient was between the size of the catch at 8 p.m. and the duration of moonlight after this period, and its value was –0·783±0·068.

4. During moonless periods, the flighting of ♀ A. funestus was partially inhibited, and when the moon came out, the waiting mosquitos entered the huts in greatest numbers. The average proportion caught when the moon rose during catching periods ending 10 p.m., midnight and 7 a.m. was approximately 50 per cent., 100 per cent. and 400 per cent. greater, respectively, than the proportion caught during these periods on moonless nights. On moonless nights many mosquitos waited until twilight of the following evening before entering the huts, and in consequence the average proportion caught in the period ending 8 p.m. was six times greater on moonless nights than on the nights after full moon.

5. The records for ♀ A. melas are less complete, and the correlation coefficient for the relation between time of entry and time of moonlight has the value 0·421±0·184, a result which is barely significant. In view of the highly significant correlations found in the case of A. funestus it is unlikely that this correlation is fortuitous.

6. The average rate of entry of ♀ A. funestus throughout one complete lunar cycle at Krabonekrom, Gold Coast, was 27 per cent. before 8 p.m., 22 between 8 p.m. and 10 p.m., 21 between 10 p.m. and midnight and 30 after midnight. The similar rates for A. gambiae (based on very few records), were 10, 20, 30 and 40 per cent. respectively. Comparable rates for A. melas in Sierra Leone are not available, but at least 45 per cent. entered before 10 p.m. (= 11 p.m. Gold Coast time).

7. The essential difference between the behaviour of ♀ A. funestus and that of A. gambiae is that the former flight earlier than the latter. A. funestus also seem to be more active at dawn, so that they are more crepuscular in their habits than A. gambiae.

8. 58 per cent, of the Anophelines caught at 8 p.m. were unfed. This indicates that early entrants usually rest after entry and before feeding, and average time of biting is later than average time of entry. Gravid or half-gravid mosquitos and those containing partly-digested blood, did not enter the huts; there was evidence that Anophelines usually became fully gorged at one feed. Of the Anophelines caught in huts after dawn 95 per cent, were fully fed.

The work described below forms part of an investigation to determine the residual toxicity of DDT when applied to the surface of materials commonly used in the construction of buildings. the bed-bug (Cimex lectularius, L.), because of its wide distribution and habit of infesting surfaces, either exposed or hidden, is an eminently suitable subject for such an investigation. In order to compare the residual toxicity of different surfaces after treatment with DDT, it was considered an essential preliminary step to ascertain whether or not the biological condition of the insect influenced its resistance to the insecticide. The present investigation was, therefore, undertaken to establish any possible variations in resistance which could be correlated with one or more of the following biological factors:— 1. starvation for different periods; 2. sex; 3. stage in the life-history; 4. temperature.

Durant l'année 1943, au cours de la campagne de lutte contre le criquet pèlerin (Schistocerca gregaria, Forsk.) il a été fait un large emploi des fruits de la pastèque (Citrullus vulgaris, Schrad.) dans les régions sahéliennes de l'Afrique Occidentale Française, des côtes de la Mauritanie à la rive occidentale du lac Tchad.

Notes are recorded on some of the habits of members of the subgenus Coquillettidia of the genus Taeniorhynchus, with an account of catching and rearing methods employed. Fourth-stage larvae of eight different species have been caught and successfully reared to the adult stage, four of these for the first time. One new species is included in the list.

Keys are given for larvae and pupae, together with full descriptions of the larvae and pupae of the eight species dealt with.

Errors in the previous descriptions of T. aurites and microannulatus are pointed out.

The practical results achieved in the past twenty-five years' work on the major insect pest problems of Mauritius are discussed.

The major insect pests studied are the white grubs of sugar-cane, Clemora smithi (= Phytalus smithi) and Oryctes tarandus; the sugar-cane moth borers, Proceras sacchariphagus, Sesamia vuteria, and Argyroploce schistaceana; the coconut scale-insect, Aspidiotus destructor; the pineapple mealybug, Pseudococcus brevipes; the cochineal insects of the prickly pears, Dactylopius tomentosus and D. indicus; fruit-flies; the biting flies, Stomoxys nigra and S. calcitrans; the red ant, Solenopsis geminata; the black ant, Technomyrmex detorquens; and termites.

The actual status of these pests is given in detail. Clemora smithi, which was first recorded in 1911, is still a major pest of sugar-cane in spite of the 43 species of parasites and 2 species of predators introduced to check it. Oryctes tarandus very seldom causes severe damage now to sugar-cane plantations. It is a sporadic pest, and has been definitely held in check since the Scoliid wasp, Scolia oryctophaga, was introduced from Madagascar in 1917.

The status of the three moth borers of sugar-cane and that of their parasites is fully dealt with. Of the five species of parasites introduced from Ceylon and Madagascar in 1938–40 to control Proceras sacchariphagus, only one has so far been recovered and its beneficial effect is very limited. An account is given of experiments on the mass liberation of the egg parasites, Trichogramma australicum and T. evanescens. The results were in both cases deceptive. The spotted borer, P. sacchariphagus, is a major pest of sugar-cane in some localities and further introductions of parasites from other cane-growing countries are recommended.

The coconut scale, Aspidiotus destructor, which, since 1927, threatened all the coconut plantations of the Island, is now under control, owing to the introduction from Ceylon and Java, respectively, of the predacious Coccinellids, Chilocorus nigritus and C. politus. The bionomics of the various parasites and predators of this scale in Mauritius are fully described.

A historical review is given of the introduction of the cochineal insects of prickly pear, Dactylopius indicus and D. tomentosus in the Island. The completely successful eradication of Opuntia tuna and O. monacantha, by these two Coccids is given as a typical example of the biological control of noxious plants by insects.

The pineapple mealybug. Pseudococcus brevipes, which is responsible for wilting disease, is considered to be one of the factors limiting the development of the pineapple industry in Mauritius. Spraying with a diesel-clay emulsion, though checking this pest, is limited in its application. Further introductions of parasites and predators are desirable.

The common white mealybug, Icerya seychellarum, is still a major pest of orchards and ornamental plants in spite of the presence of the local Coccinellid predators, Rodolia chertnesina and Exochomus laeviusculus. Practical control can be achieved by the use of banding grease coupled with the attacks of predators. No local parasite has been recorded in Mauritius.

Six species of Trypetids are recorded. Of these, the flies attacking cucurbitaceous fruits are the most harmful. Experiments with numerous attractants for fruit-flies are mentioned. The only bait showing any success in the control of the Dacus flies is the Mally fruit-fly bait. Parasites have been imported from South Africa and Ceylon without success. Further attempts should be made.

The bionomics and breeding places of the blood-sucking fly, Stomoxys nigra, are described.

The Red Ant, Solenopsis geminata, is a major pest of seedlings in Mauritius. It is now destroyed by the use of a diesel-clay emulsion. Baits containing thallium sulphate have been tried without success.

The rôle of the Black Ant, Technomyrmex detorquens, in preventing development of the Coccinellid predators on Icerya seychellarum is emphasised. Barber's poisoned bait has been found very useful against this household pest.

A review of the species of termites found in Mauritius as well as the various means of control used against them are given. An insecticidal powder containing arsenious oxide has given the best results.

During the last 25–30 years, 64 species of parasites and predators have been imported for the biological control of eight species of insects of major economic importance and of two species of noxious plants.

The results achieved indicate that the introduction of parasites and predators in insular countries like Mauritius should be carried out only after a thorough investigation of the various ecological, biotic and climatic factors existing between the country of exportation and the country of importation.

The rôle of cyclonic disturbance in the fluctuation of the parasite population is mentioned. Scoliid parasites, for instance, are generally scarce after violently disturbed weather and may sometimes remain so for two or three years after the disturbance.

1. The danger of the establishment of yellow fever in certain of the forests of Kenya is pointed out. Kenya has had two recognised cases of the disease in recent years.

2. The Kaimosi forest was selected as a suitable area for investigation and the general environmental conditions are described.

3. A mosquito survey was undertaken over a period of two years and particular attention was paid to the question of “acrodendrophily” i.e., the preference shown by certain species of mosquitos for life in the tree-tops.

4. Five species of mosquitos—Culex nebulosus, Aëdes capensis, A. fraseri, A. deboeri subsp. demeilloni and A. longipalpis breed in high tree-holes; the last, almost entirely so. The many characteristically low breeders include A. aegypti and A. africanus. There is an interesting distinction between the two commoner Megarhines; M. brevipalpis breeds high, M. barbipes, low.

5. Four species of mosquitos—A. apicoargenteus, A. fraseri, A. africanus and A. longipalpis were found biting in the tree-tops; the last two being rare in ground catches.

6. Forest with a broken canopy proved best for adult catching and the morning was a better time than the (usually overcast) afternoon. Night searches gave negative results. Catches on high tree platforms appeared to be more prolific than those oa the ground.

7. The following points of special interest in regard to individual species of mosquitos are summarised below: —

(i) Megarhine larvae consume about 20 fourth-stage aegypti larvae a night. Pupal stage lasts on an average for nine days.

(ii) Aëdes aegypti breeds commonly in two sites (a) recently felled trees at forest edge; (b) rock pools in rivers in drought periods. Larvae are found a mile or more inside the forest. Breeding of this species has been discovered as high as 6,500 feet.

(iii) Aëdes apicoargenteus was the commonest adult. Some specimens survived 55 days in captivity.

(iv) Aëdes africanus shows a seasonal incidence, being commonest at the end of the “long rains”. Its favoured breeding places are the buttresses of the tree, Craibia elliottii.

(v) In the larval stage, Aëdes capensis is the commonest mosquito.

(vi) The characteristic breeding place of Theobaldia fraseri contained 79 parts of oxidisable matter per 100,000 of water.

8. Probably at least two trees per acre have rot holes with larvae and about 14 per cent, of the Dracaena plants are infested. Holes in shaded granite boulders are another common source of sylvan mosquitos.

9. Surveys of adjacent forest areas at different altitudes revealed striking changes in the mosquito fauna. Aëdes pulchrithorax is the prevalent species at the 6,500 foot level; A. kapretwae at the 8,500 foot; whilst in general, A. africanus, A. aegypti and A. bambusae appear to be widespread.

10. A study of the lists of mammals, trees and mosquitos indicates that the relict forest of Kaimosi contains numerous “western” species which are not found elsewhere so far to the east.

11. Immunity tests performed at the Yellow Fever Research Institute, Entebbe, demonstrated the absence of yellow fever in man and monkeys at Kaimosi. The monkeys show, however, immunity to Semliki forest virus and either Japanese B or St. Louis encephalitis virus.

12. The human inhabitants suffer from malaria and onchocerciasis (the latter carried by Simulium neavei) and the monkeys and fruit bats are infected with species of Plasmodium.

13. The Kaimosi forest from every standpoint (including meteorological) is suitable for the establishment of jungle yellow fever and possible measures of control ate briefly suggested.

14. Short descriptions of new species, etc., of mosquitos are given in the Appendix

The sandflies of Cyprus have been examined. The following ten species were found: P. papatasii, P. perniciosus var. tobbi, P. chinensis (a local race), P. perfiliewi (rare), P. sergenti, P. alexandri, P. larroussei (rare), P. fallax cypriotica, var. n., P. azizi, sp. n., and P. parroti.

Melanic forms of P. parroti were found. Melanism was associated with a reduction in the teeth of the pharynx in the female in the few specimens collected.

The race of P. chinensis found on the island differs from those hitherto described.

The general composition of the sandfly population of Cyprus is different from that of any other part of the Mediterranean examined.

P. major which is common in Palestine, Syria and the Balkans was not found.

Canine kala-azar is common on the island. P. perniciosus var. tobbi and a race of P. chinensis should be considered as possible carriers but this point must be determined experimentally.