Adults of two subspecies of Entypotrachelus meyeri Kolbe defoliate tea established on land cleared of montane rain forest, E. m. meyeri Kolbe occuring in Kenya and Tanzania to the west of the Great Rift Valley and E. m. micans Hust. only in the foothills of Mt. Kenya east of the Rift. Both subspecies are polyphagous and the host-plants are listed. Oviposition by E. m. meyeri occurs throughout the year; eggs hatch in 10–29 days, and the larvae feed on the roots of tea and various woody plants. The adults cannot fly, but field observations showed they could disperse 2–23 m in 14 days; mark-recapture data showed that 50% of the females lived 32 days, and males 37 days. The bionomics of E. m. micans were similar, but adults lived longer (50% of the females lived 55 days, males 49 days). The leaf area consumed by adults of E. m. micans was 66 mm2/day whereas the total leaf area of a seedling is 30–50 cm2, and it is shown that three adults per plant could effect defoliation in 21 days, assuming no new leaves are formed; leaf destruction at this rate is not considered economically important. Larvae of E. m. micans caused 37% rejection of tea seedlings through destruction of the tap root.

A technique is described for the preparation of large quantities (up to 1100 kg) of plain and waterproofed leaf-cutting ant bait from dried citrus meal, soyabean oil and aldrin. Small amounts (500 g) were weathered in sun and shade during wet and dry periods in Trinidad and the acceptability of the bait to ants {Acromyrmex octospinosus (Reich) and Atta cephalotes (L.)) and the persistence of toxicants in it were assessed in laboratory tests. Methyltrichlorosilane was the most effective of three waterproofing agents tested. In the wet season, waterproofed bait remained acceptable to ants for at least 30 days whereas plain bait deteriorated in less than four days; in the dry season waterproofed bait was only marginally more attractive. In the wet season, Acromyrmex strongly preferred weathered waterproofed bait to fresh leaves, whereas Atta slightly preferred leaves to bait. In the dry season, Acromyrmex showed little preference between bait and leaves, whereas Atta strongly preferred bait. Over the period that the bait remained attractive to the ants it retained between one-third and two-thirds of the original aldrin in the form of aldrin and dieldrin.

Operated in sheltered woodland, the samples obtained by Rothamsted tungsten-filament and Robinson mercury-vapour traps are best described by a spatial model different from that used in a site exposed to wind. Differences between the proportions of Noctuidae and Geometridae were accountable to differences in height of flight. The Rothamsted traps gave more consistent samples than the Robinson traps and averaged about onequarter the number of moths.

A species of Mompha, discovered on creeping water-primrose Ludwigia adscendens (Onagraceae) in Assam, N.E. India, is described as new and named Mompha ludwigiae. The eggs, larval and pupal stages are also described. Females lay up to 150 eggs in 3–7 days; development from egg to adult takes 3–5 weeks, the larva at first mining the leaves and later feeding externally on terminal shoots. The larvae could be reared on Oenothera rosea, L. perennis and L. octovalvis in the laboratory, but no host-plants other than L. adscendens were found in the field. A Eulophid parasite, Euryscotolinx sp., has been bred from pupal cocoons.

Studies on mosquitoes of the Anopheles gambiae Giles complex at Bwamba, Uganda, showed that the newly recognized sibling species D may be restricted to breeding areas where waters contain unusual solutes, the effluent from hot springs at Buranga being particularly favourable. Two other members of the complex, species A and B, were also detected in the neighbourhood apparently breeding only in more normal situations. Species D specimens were usually identified cytotaxonomically, from the presence of XR potytene chromosomes of the type also occurring in species C and A. melas (Theo.) (neither of which is known from the area). Most species D females have distinctively pale palps, a valuable although limited distinguishing feature. Features of setation on larvae and pupae were discerned which might serve usefully as meristic chaetotaxonomic factors, at least locally, if tested further. Although it is unquestionable that a large proportion of species D females forage in the Semliki forest, feeding on primates, bovids, suids or unsuspected hosts, it was demonstrated that many species D females are markedly anthropophilic and display strong endophilic and endophagic tendencies, to the extent that they may be primary vectors of human malaria and filariasis over a limited range. Species D males were rare indoors and remained mainly within the forest, where great numbers rest low on vegetation in company with unfed females, near to breeding areas of brackish marsh. Two polymorphic chromosomal inversions in species D apparently embody the genetical causes of its differentiation. Cytogenetical evidence indicates that species D is phylogenetically nearly intermediate between A. melas and species C of the A. gambiae complex.

Life-cycle variation, involving alternative methods of overwintering (holocycly and anholocycly), is found in many important pest aphids, and may have considerable ecological, genetic and economic significance. The occurrence of life-cycle variation in Myzus persicae (Sulz.) in different parts of the world is reviewed. The holocycle of M. persicae, with sexual reproduction and overwintering of eggs on Prunus, occurs in the temperate regions of every continent, and although anholocycly is widespread in warm climates there are indications that the potential for sexual reproduction may be retained throughout the whole range of the species. The environmental factors involved in the determination of sexual morphs are considered. Photoperiodism provides the timing mechanism for sexual morph production, and the photoperiodic response is probably tuned to different regional conditions and buffered against other environmental variables within certain limits. The only other environmental factor clearly implicated in sexual morph production is temperature, and an attempt is made to correlate the global pattern of life-cycle variation in M. persicae with temperature differences between regions, by dividing the world into six zones delimited by selected isotherms. The overall pattern of life-cycle variation can be explained largely on the basis of climatic differences between zones, and the effects of these differences on gene and genotype frequencies within the species. Clones of androcyclic character seem to be widespread and, although genetically isolated anholocyclic biotypes undoubtedly exist, androcyclic clones could be the principal source of parthenogenetic overwinterers in all parts of the world where life-cycle differences exist within populations.

Aircraft were used to apply a bait containing aldrin, soyabean oil and citrus meal against leaf-cutting ants (Acromyrmex octospinosus (Reich) and Atta cephalotes (L.)) in Trinidad. The distribution patterns of bait applied by different aircraft were measured. Plain bait applied at 2.2 kg/ha to an uncultivated island in the dry season destroyed 91% of nests of Acromyrmex, and waterproofed bait applied in the wet season to cultivated land on the mainland destroyed 85%. Small nests of Atta were also destroyed by this treatment but large nests in forest required much heavier dosages applied to the nest itself. Lizards and crabs living in the baited areas were contaminated with aldrin but there was no evidence to show that they were harmed.

The pattern of spontaneous activity of the tsetse fly Glossina morsitans Westw. was measured in LD 12:12 at constant low temperature. At 22°C the typical V-shaped pattern was unchanged; at 19°C, when activity was reduced to five flights per day, the morning arm of the V was retained but the evening one suppressed. Cold therefore changes the shape of the endogenous rhythm only slightly, and its effects in the field must concern activity level rather than pattern. Examination of the published data shows that the correlation of field activity with temperature is not as good as is generally supposed.

In this method the toxicant is absorbed with blood serum onto a loose roll of chromatography paper; first-instar larvae of Lucilia cuprina (Wied.) are inserted into the air spaces of the roll and allowed to feed. Mortality of the larvae is assessed 24 h later. Typical results using diazinon, dichlofenthian and chlorfenvinphos against susceptible and resistant strains of L. cuprina showed larvae to have resistance factors 6–8 times higher than those for adults determined by established techniques.

Third-stage larvae of the following species of Dynastidae associated with coconut palms are described: Xylotrupes gideon (L.) (including four subspecies, X. g. ulysses (Guérin), X. g. szekessyi Endrődi, X. g. beckeri Schaufuss and X. g. asperulus Minck), Scapanes australis (Boisduval) (including three subspecies, S. a. grossepunctatus Sternberg, S. a. salomonensis Sternberg and S. a. australis (Boisduval)), Trichogomphus fairmairei Arrow, Oryctes rhinoceros (L.), O. centaurus Sternberg, Oryctoderus latitarsis Boisduval and Oryctoderus sp. (either O. coronatus Bates or O. godeffroyi (Fairmaire). A key to species, based on larval characters, is included. Some small differences, particularly in setation, were found among the four subspecies of X. gideon examined, and differences in setation of the abdominal segments were found between S. a. grossepunctatus and S. a. salomonesis on the one hand, and S. a. australis on the other.

Two toxic baits, one containing aldrin and the other containing mirex, were compared for their effectiveness in killing field colonies of Atta cephalotes (L.) and Acromyrmex octospinosus (Reich) in Trinidad, West Indies. The fate of these two toxicants within tie nests was followed by using radiolabelled bait. Of the two baits, mirex appeared to be more acceptable to both species of ant. With Acromyrmex, the ratio of living to dead fungus gardens was not affected by the bait used, whereas with Atta there was a significant difference, aldrin killing many more gardens than mirex. For both species the proportion of medium-sized workers radiolabelled was significantly greater than for the other castes, and mirex marked greater numbers of all castes than did aldrin. Greatly different proportions of Atta workers were radiolabelled in the different fungus gardens of a nest with both baits. More radioactivity was detected in the fungus gardens treated with mirex, the difference being significant for Atta but not for Acromyrmex. Mirex bait particles incorporated into the fungus gardens were significantly smaller than aldrin bait particles, indicating different degrees of comminution by the ants.

The losses caused by leaf-cutting ants (Acromyrmex octospinosus (Reich) and Atta cephalotes (L.)) to citrus and cocoa crops in Trinidad were assessed as about $40 000 TT per annum in each of these crops. Losses to other crops were unassessable on a national scale but often important to individual farmers. Aerial baiting, using a locally produced bait of aldrin, soyabean oil and citrus meal, would be at least as, and probably more, effective than current on-farm control methods. The cost was $6.18/ha, approximately four times cheaper than current methods and twice as cheap as searching for nests and applying the same bait to them by hand. Hand baiting with the local bait would be almost twice as cheap as using the imported alternative available. At the dosages and frequency of application recommended (2.2 kg of bait containing 0.4% technical aldrin/ha/annum) the total amount of organochlorine insecticide currently used against leaf-cutting ants in Trinidad would decrease between two- and ten-fold, depending on the area treated. Serious long term residual effects are most unlikely to develop, and if bait were applied carefully no harmful effects to man or stock would occur. Farms larger than four hectares would benefit proportionately more than smaller farms, unless charges were manipulated. Fewer men would be needed to control ants than at present. Losses to traders caused by a decrease in the sales of alternative insecticides would be trivial. A national dry-season baiting scheme would utilize more fully aircraft already available in Trinidad, and would help export crops and the production of home-grown foods.

A method for collecting flying insects is described, developed for studies on the dispersal of Simuliidae. The apparatus consists of a net mounted on the roof of a motor vehicle, leading to a catch-segregating apparatus within it, such that the catches over any desired distance at speeds up to 48 km/h can be segregated without stopping until 50 catches have been obtained. The volume of air filtered is measured with a counter anemometer mounted in the net mouth. Specimen results for one day are given as an illustration of the efficiency of the method.

The management of the sugar-cane crop in Trinidad involves widespread aerial spraying against the sugar-cane froghopper (Aeneolamia varia saccharina (Dist.)). For control decisions to be made on an economic basis, information concerning the damage and control relationships associated with this pest is required. Assuming a linear damage function, a cost-effective analysis based on the relative cost of eliminating adult-days is made concerning the choice of insecticide and the decision—when to spray. It appears that a first-brood spraying strategy is more cost-effective than a strategy of later-brood spraying alone, provided that the density and duration of the first-brood can be accurately monitored.

In laboratory experiments with Prays citri (Mill.) in Israel, females 1–16 h old proved more attractive to 2–6-day-old males than did female pupae 1–8 h before emergence or female adults 1–4-days old. Attractiveness in the field was tested by means of sticky traps baited with female pupae or adult females. Females were attractive for up to 14 days in summer (19–30°C) and 27 days in winter (8–20°C). Greatest attraction was shown by 1–7-day-old females during summer as compared with 5–9-day-old females during winter. Field experiments with baited sticky traps attached to wind vanes showed that more males were caught by traps aligned with the wind than by those placed at right-angles to it or hung among tree branches, by traps 1–7 m above ground than those 30–50 cm above, and by traps five metres distant from trees than those 10–80 m distant.

Larvae of Aphidoletes Kieffer and Monobremia Kieffer feed exclusively as predators on aphids. The taxonomic status of these genera and their included species is reviewed and new generic and specific synonymies are given. A. aphidimyza (Rondani), A. urticariae (Kieffer), A. abietis (Kieffer), A. thompsoni Möhn and M. subterranea (Kieffer) are accepted as good species and a neotype is designated for A. urticariae. Diagnostic characters of these species are indicated and characters of the male genitalia, female wings and larval skins are illustrated. Published information on their biology, particularly of the commonest species A. aphidimyza, is summarized and includes details of host ranges and geographical distribution. Lestodiplosis grassator (Fyles), L. pini Barnes, Trilobia aphidisuga Del Guercio, Uncinulella eriosomiperda Del Guercio, Trilobiella siphae Del Guercio and Adelgimyza strobilobii Del Guercio probably prey on aphids but have not been recognized since they were originally described. Comparative assessments of field populations of Aphidoletes by examination of aphid colonies in the field, by incubation of samples in polythene bags and by laboratory examination of samples fixed in 70% ethyl alcohol showed that only the last gives accurate counts of eggs and larvae. Counts of aphids, Aphidoletes, Syrphids, Coccinellids and Anthocorids in 35 samples taken from aphid colonies occurring at Wisley in 1970–72 are tabulated, together with seven samples from other localities. Aphidoletes (mainly A. aphidimyza) was numerically the dominant predator in all three years at Wisley and was about ten times more numerous on wild plants than on cultivated plants.

An account is presented of the distribution and amounts of moonlight in latitudes near the equator. This includes a Table on the amount of moonlight for each hour of the night throughout a standard lunar cycle, applicable to any locality between 10°N and 10°S, and a Table of standard groups of moon phase which can be used at any locality irrespective of latitude. The construction of these Tables is described in detail and their use briefly discussed. A method is described which enables light-trap catch records to be arranged for analysis directly against moon phase.

Achrysocharis diplosidis reared from galls of the Cecidomyiid Hyperdiplosis eupatorii Felt in Trinidad and Brazil is described, A. pulcherrima Kerrich is recorded from Pakistan, and a key separating these two gall-associated species of Achrysocharis Girault is given. Larval or pupal parasites of Promecotheca cumingii Baly (Hispidae) in Sri Lanka are described as Cotterellia lankensis, and of Promecotheca spp. in New Britain, Admiralty Is. and New Guinea Territory as C. insularis; C. podagrica Waterston is redescribed, and a key is given to the three species of Cotterellia Waterston ; the generic relatives of Cotterellia are discussed and a key separating this genus from Platocharis Kerrich and Schizocharis Kerrich is given. Rekabia Cameron is placed in synonymy with Euplectrus Westwood, and notes are given on material for a classification of the relatives of Stenomesius Westwood.

Weekly catches of five species of mosquitoes over nearly two years at Belém, Brazil, were plotted against accumulated rainfall during the four weeks before catching. Rain-pool-breeding species (Aedes serratus (Theo.), A. scapularis (Rond.), Psorophora ferox (Humb.)) show a marked peak in catch size (population density) at 100–180 mm accumulated rainfall; one swamp-breeding species (Culex portesi Senevet & Abonnenc) shows no relation to rainfall while another (C. taeniopus D. & K.) reaches a plateau at 180 mm. A more detailed analysis of catches in relation to rainfall distribution showed that maximal populations of A. serratus appear 16 days after 39.7–87.9 mm of rainfall has accumulated over 14 consecutive days, which should include 3–7 dry days; P. ferox requires 55.2–99.4 mm rainfall in the same period; and A. scapularis requires 26.5–114.9 mm rainfall, with a marked alternation of wet and dry days.

To determine whether the brain medial neurosecretory cells (MNC) provide a hormone regulating ovulation and larviposition in females of Glossina austeni Newst., the effects of MNC ablations, sham operations, MNC implantations, and MNC extract injections were investigated. Ablation almost always inhibited ovulation and interfered with larval development and larviposition; it did not appreciably influence blood feeding, digestion or egg maturation. However, surgical brain trauma and even general stress also inhibited ovulation in control flies to some extent, though inhibition was more frequent and prolonged after greater trauma or incisions closer to MNC. A similar relationship existed between proximity of incisions to MNC and success of larval development and larviposition. MNC implants and extracts failed to induce ovulation in ablated flies. Most evidence indicates that MNC ablation affects ovulation and larviposition, not by removing a source of stimulatory hormone, but by interfering with nervous pathways necessary either for hormone release at some other site or for direct stimulation of the reproductive tract musculature.