The reasons for adopting the name “ leaf-curl ” for the virus disease which is manifested by symptoms of “ leaf-crinkle ” on certain varieties of cotton, and by “ mosaic ” on certain others, are discussed.

A brief review is given of the very few references in previous literature which are relevant to leaf-curl.

Up to the end of February 1931, 168 straightforward attempts had been made to transmit crinkle from Sakel cotton to Sakel by means of white-flies (Bemisia gossypiperda, Misra & Lamba), of which no fewer than 157 were successful. No transmission was obtained with any other insects.

A single white-fly can transmit the disease to a healthy plant, though infection is less regularly obtained when one or only a few white-flies are used.

There is a large amount of negative evidence which indicates that the virus is not transmitted through the seed of Sakel cotton. There is also no evidence that it can be transmitted through the soil.

In nearly 200 controlled experiments, the incubation period of the virus in the plant varied from 8 to 34 days, but over two-thirds of the recorded periods were between 11 and 19 days. Possible causes of the variation in the incubation period are discussed.

Little is known with certainty about the factors which influence the severity of the disease.

At least one case of complete recovery from the disease appears to have been observed. Partial recovery is not infrequent.

During July and part of August 1929 an experiment was carried out at Sambala (Western Kondoa), Tanganyika Territory, designed to show that : (i) Hungrier flies marked in Berlinia- Brachystegia woodland left that type for the supposed feeding-grounds (“ female centres ”) earlier than repleter flies ; (ii) conversely, repleter flies remained within the woodland longer than hungrier flies ; (iii) hungrier flies in the supposed feeding-grounds remained in them for the first few days ; and lastly (iv) repleter flies in these situations returned to the Berlinia-Brachystegia woodland to remain there until later.

The present paper is the result of studies begun privately by the writer shortly after his return from South America in February 1930. The subsequent continuation and conclusion of the task was made possible through the authorities of the Imperial Institute of Entomology, to whom, especially the Director, Sir Guy A. K. Marshall, C.M.G., F.R.S., the writer is gratefully indebted for his appointment, in a temporary capacity, at the Farnham House Laboratory, as from 1st August. It was considered that a final revision of the American forms of the groups here considered would be a useful contribution to the West Indian investigations on sugar-cane insects being conducted on behalf of the Institute by Dr. J. G. Myers. At the laboratory the writer found conditions ideal for the prosecution of such work, and he desires to acknowledge his appreciation of the continuous help and encouragement received from the Superintendent, Dr. W. R. Thompson, and his colleagues there, among whom Dr. G. Salt deserves a special word of thanks for his never-failing generosity in offering friendly advice and practical aid, both of which, it must be confessed, have been frequently needed.

In the case of silica, there appears to be a definite variation of SiO2 content with different species. The difference between yellow locusts (6·27 per cent.) and red locusts (14·7 per cent.) is especially marked. It may be attributed to the fact that red locusts are actively feeding, sexually immature adults, and in consequence they consume large amounts of food with a high silica content. The fact that museum specimens of yellow locusts contain a higher percentage of silica than intact yellow specimens indicates that the majority of the silica is not concentrated in the digestive tract. The same appears to apply to the iron content. Copper was found in appreciable quantities in all the insects examined, but no definite conclusion can be drawn from the numerical results, as the variation from batch to batch was too great. The results for manganese are not satisfactory, as in several cases negative evidence was obtained. This is inconsistent with the work of Vinogradov, who obtained conclusive evidence of manganese in insects. In the case of Agonoscelis, however, the manganese content was found to be extraordinarily high (0·47 per cent.), suggesting that it may play a definite physiological part. The presence of a respiratory pigment, pinnaglobin, containing manganese, has been shown in the tropical mussel, but there is as yet no evidence for the presence of this pigment in the insect world.

It is impossible, however, to draw any such conclusions from ash analysis results of entire insects only, and systematic analyses of the blood of numerous species will be necessary to establish the presence of the metals iron, copper and manganese, in their respiratory pigments.

Definite evidence of nickel (in almost negligible amounts) was shown by the large-scale analysis, but apart from this no element was detected which had not been previously reported.

The present paper attempts to summarise the information collected by the Imperial Institute of Entomology on the swarming of migratorioides in Africa.

It is concluded that the outbreaks in West, Central and East Africa are all interrelated and probably originated in one or at most two permanent breeding-grounds.

The area most suspected is the lake district of the middle Niger, where conditions seem suitable for the formation of the swarming phase. Lake Chad is probably a secondary rather than a permanent breeding-ground.

It is indicated that the key to the problem of the cause of the original outbreak may be found in the study of recent variations in the water régime of the River Niger.

Should these suppositions be substantiated, we are well on the way towards a solution of the problem. The ultimate control of the locusts depends not on control during the swarming periods, but upon preventing the swarming phase from developing in the permanent breeding-grounds.

It is therefore important to check and augment the records of the early swarms in West Africa and more important still, to study the behaviour of the locust, its ecology, and the water régime in the suspected areas.

In conclusion the author wishes to acknowledge the constant help and advice of Mr. B. P. Uvarov. His thanks are also due to Mr. H. B. Johnston for his permission to make use of his paper in manuscript.

Beattie & Howland's recent studies of the bionomics of some tree-hole mosquitos emphasize the importance of the food factor, total organic nitrogen, and the apparent lack of correlation between hydrogen ion concentration and larval incidence. In Queensland mosquitos selecting such breeding-places are with a few exceptions distinctly arboreal, and the presence of domestic mosquitos in tree-cavities is dependent upon proximity to human habitations. It is a curious fact that when this is the case both Aëdes argenteus and Culex fatigans will breed in smaller accumulations of water than Aëdes notoscriptus does.

Appended herewith are the results of a recent survey made in midsummer of trees situated within the Brisbane Area of 385 square miles. Pine trees and gum trees do not lend themselves to the production of cavities and therefore offer very little opportunity for breeding.

The poincianas hold the highest actual percentage, but on account of the far greater number of figs containing potential breeding cavities, the latter constitute a greater actual menace. The amount of water contained in each cavity varies considerably with the time of year.

The larvae of Uranotaenia ornata, Theo., are commonly found in pineapple plants in the neighbourhood of Lagos, Nigeria, and invariably in company with Harpagomyia larvae.

During the course of certain investigations, the writer had occasion to collect a considerable number of mosquitoes in Nigeria, particularly in the vicinity of Lagos. Several species of Uranotaenia were represented in many of the collections. In view of the difficulties experienced in identifying certain specimens, opportunity was taken during a recent visit to the British Museum (Natural History) to study the African Uranotaenia in the collections there. Subsequent to the publication by Edwards in 1912 of the last generic treatment and key for the separation of the species, several additional species have been added to the list from the Ethiopian region, and more materials have accumulated, thus facilitating further study and permitting a reconsideration of certain specific assignments.

Erythroneura lubiae, sp. n.

Colour :—Pale green covered with a whitish pruinosity ; middle of pronotum and apices of tibiae, tarsi and ovipositor of ♀, slightly darker green ; a spot at middle of apical margin of pronotum and a broad stripe down middle two-thirds of scutellum, opaque white. Eyes dark greenish grey, apex of rostrum brown. In some female specimens the head and pronotum are uniformly greenish yellow and the white spots are obsolete. Tegmina more or less hyaline, the veins, clavus, brachial and suprabrachial areas, sub-opaque, suffused with pale green or greenish yellow ; the membrane infuscate towards apex and along inner margin ; claval commissure very narrowly and extreme inner margin of membrane beyond apex of clavus, dark brown ; costal plaque distinct, pallid. Wings whitish hyaline, faintly infuscate towards apex, veins white. Abdomen pale whitish green, basal tergite and anal segment greenish yellow. Tibial spines white.

Structure :—Ocelli absent ; head as wide across eyes as pronotum across humeral angles (47 : 47) ; seen from above slightly less than half as long in middle as wide between eyes (13 : 27), only slightly longer than length at eye (13 : 11) ; face (including clypeus) about 2·25 times as long as wide between antennae (50 : 22) ; frons three times as long as clypeus (36 : 12), which is apically sub-angulate. Pronotum about half as long as wide across humeral angles (23 : 47).

Tomaspis saccharina, Dist., is a Cercopid bug which infests the cane-fields of Trinidad in vast numbers during the wet months of the year. Corresponding with the successive generations, there are three, and sometimes four nights or broods of adults, each of which causes severe damage to the cane. This damage, or “ froghopper blight,” as it is locally called, is a result of the feeding activities of the adult on the foliage, and is characterised by a progressive discoloration of the leaves, leading to death of the affected leaf and, in severe cases, to the death of the cane. In either event, growth of the plant is more or less severely checked and the adverse effect on yield is usually of a very significant nature.

Specimens of the two Coreid bugs described below were submitted by the Entomological Division of this Department to the Imperial Institute of Entomology for determination and were returned shortly afterwards with the intimation that both were new to science.

The pseudopupa and last larval instar of Epicauta erythrocephala, Pall., have remained undescribed up to the present, although the economic importance of the insect is very great, the beetle being an important pest of many cultivated plants and its larva a parasite of locust egg-pods. These stages are, however, figured in the work of I. Portchinsky, published in 1914, but in the figure of the pseudopupa the stigmata on the three thoracic segments are omitted.

During investigations made in early spring (1926–1927) in the environment of Old Bukhara, I found a great quantity of the pseudopupae and larvae of the last instar of E. erythrocephala in the soil. In the early spring of 1929 the same instars were reared in the insectary of the Agricultural Experimental Station at Shirabudin (Old Bukhara).

The larval stages of two species of Syrphid flies, namely, Eumerus tuberculatus and E. strigatus, are frequently found in rotting narcissus bulbs, in iris and tulip, and in vegetable roots, such as parsnip. The writer has shown that frequently the decay, at least in the narcissus, is directly attributable to the presence of these larvae and that they must therefore be considered to be primary pests. It was further shown that, contrary to current opinion, the predominant species at the time of writing was E. tuberculatus. In passing it may be recorded that this predominance has, in each subsequent year, become more marked, until at the present time E. strigatus has ceased to rank as a bulb pest, at least in South-West England, Holland, and probably the Irish Free State. The larvae of the two species are very similar to one another, and it has not been found possible to separate the species, with any great degree of accuracy, during the immature stages. In the adult form, as shown by Collin, separation is easily effected. In the course of experimental work with these flies a third, and superficially somewhat similar Syrphid, namely Syritta pipiens, L., was frequently bred from decayed narcissus bulbs.

In a recent paper by the writer (Bull. Ent. Res., xxi, 1930, pp. 201–256) it was shown that the apparent fly density varied according to season. It has now been found that these seasonal variations in the fly density can be correlated with the evaporation rate.

The accompanying graph shows the mean monthly evaporation rate, and the mean monthly apparent fly density. The evaporation rate figures have been obtained from a Livingston atmometer at the Kikori Entomological Station. The fly figures have been taken from the North-East Kikori Round, because this fly-round is not subjected to large game movements and hence is considered to be the best for demonstrating the correlation between fly density and season.

The graph deals with a period of eighteen months, and it is clear that a relationship does exist between the two curves.

In June and July 1929, the fly density was at its maximum. Hence it is reasonable to postulate that, at this time, the evaporation rate was suited to the requirements of the fly community.

The following notes deal with a collection of sandflies from Africa sent for identification by Sir Guy Marshall, C.M.G., F.R.S., of the Imperial Institute of Entomology, by Dr. Pontes from Mozambique, by Major W. F. M. Loughnan, R.A.M.C., Mauritius, and with a number of specimens lent to Prof. Adler by the London School of Tropical Medicine. I am indebted to Prof. Adler for placing this material at my disposal.

Among the insects which do serious damage to grasses grown for seed, the larvae of gall midges (Cecidomyidae) are of great importance. Very little is known about their bionomics, except in the case of the Sorghum midge (Contarinia sorghicola), and the Meadow Foxtail midges (Dasyneura alopecuri, Contarinia merceri, and Stenodiplosis geniculati). In view of this and the fact that several grasses are grown for seed with advantage in Great Britain, it has been thought advisable to collect together all the available information.

Through the courtesy of the Imperial Institute of Entomology, I have been able to examine specimens of a remarkable gall midge reared by Mr. E. Hargreaves, in Sierra Leone, from larvae predacious on a species of mealy-bug.

For a number of years the western grass-stem sawfly, Cephus cinctus, Norton, has been accounted one of the major insect pests of wheat in Western Canada. Originally a grass-feeder, it began to attack wheat about the beginning of the present century, at first only occasionally and at the edges of fields in Manitoba, later much more commonly and with a very wide distribution throughout the wheat-growing areas of Manitoba, Saskatchewan, and Alberta. While this sawfly was confined to native grasses it was held in check by two principal factors (Criddle, 1922) : shortage of grass stalks suitable for oviposition, and parasites. When, however, it began to attack wheat, both of these restraints were evaded. In the first place the wheat fields provided an abundance of host-plants. In the second place, the natural parasites of the sawfly did not move with it into the grain fields—whether they are unaware of the presence of their host in wheat plants or are unable to attack it in that situation is unknown.

The following data have been gathered over a period of about one year (January to December 1929) in the Nairobi District, Kenya Colony.

It was thought that the organic content of natural waters would certainly offer some indication of their suitability for the breeding of Anopheline mosquitos and that an investigation might suggest some definite line of approach to this intensely complex problem.

During November and December 1928, certain water sources were selected by one of us (C. B. S.), in widely separated parts of the district, for weekly sampling. In each separate area some four to eight places were chosen for routine searches for Anopheline larvae ; thus in the Q area (Table II) seven places were searched every week and in the B area (Table I) six places. From the places in each area two were chosen for routine oxygen absorption tests, one with Anopheline larvae and one without, water from these same two places being sampled and tested weekly. Thus in the Q area, spots Q 5 and Q 6 were used for this purpose, and in the B area B 3 and B 5. It is necessary to emphasise the fact that these places were separated from one another by distances of 50 to 150 yards. The choice of the source of water to be sampled was made purely on the presence or absence of Anopheline larvae at the beginning of the work.

The following brief contribution summarises a part of the contents of an extensive Russian report on parasites of the egg-pods of some Turkestan Acridodea, studied by the writer while on the staff of the Middle-Asian Institute of Plant Protection (Tashkent, U.S.S.R.), in the years 1927–30.

The writer is specially indebted to Mr. V. I. Plotnikov, Director of the Institute, under whose supervision the work was done, and who kindly identified the Acridian hosts ; and for the identification of the parasites reared, to Messrs. A. A. Ogloblin, Santa-Anna (Scelionidae), S. J. Paramonov, Kiev (Bombyliidae), B. B. Rohdendorf, Leningrad (Tachinidae) and B. S. Kuzin, Moscow (Meloidae). The author's thanks are due also to Mr. B. P. Uvarov, for his kind assistance in the publication of this note.