The distribution of ants in a 2·25-ha cocoa plantation at the University of Ghana Agricultural Research Station, Kade, was recorded in December 1970. The 44 species of ants found could be grouped as dominant, co-dominant, sub-dominant and nondominant. Oecophylla longinoda (Latr.) and Macromischoides aculeatus (Mayr) were mutually exclusive as were six Crematogaster spp.; the latter were also exclusive of O. longinoda and M. aculeatus. Camponotus acvapimensis Mayr became dominant under insolated conditions, and Platythyrea frontalis Emery for reasons unknown. The distribution of these ant species in cocoa plantations resembles a three-dimensional mosaic which is preserved by their aggressiveness and by competition for food and nesting sites. It is suggested that this mosaic is due to vegetative succession.

Dentichasmias busseolae Heinr., an important pupal parasite of Pyralid graminaceous stem-borers in East Africa, is widely distributed in the Ethiopian Region but is more abundant in areas with an equatorial humid climate. In the laboratory adults mated immediately after emergence and started oviposition the same day. The female oviposits only in a borer pupa without a cocoon in a stem, and oviposition is stimulated by the presence of the frass of Chilo partellus (Swinh.) larvae. D. busseolae does not attack Busseola fusca (Fuller), Sesamia calamistis Hmps. or Eldana saccharina Wlk. in the field but these were accepted for oviposition when exposed in C. partellus pupation tunnels or with frass of C. partellus larvae, and were suitable for development. D. busseolae could complete its development in one- to eight-day-old C. partellus pupae when the pupal period of C. partellus was 10·5 days. The life-cycle was completed in 15–19 days at a mean temperature of 25°C, but could not be completed at a constant temperature of 30°C; temperatures fluctuating between 22°C and 30°C were not detrimental. At Namulonge (Uganda) percentage parasitism showed a positive correlation with rainfall, with a lag of about a month.

Observations on Aphis craccivora Koch at Serere, Uganda, in 1964–68 showed that dry season carry-over of aphids could occur on the common weeds Euphorbia hirta and E. prostrata; at other times aphids were seen on Gliricidia sepium, Cassia tora, Centrosema pubescens and Macroptilium atropurpureum, the last two being increasingly used as pasture legumes. In pot tests, attempts to transfer the virus through G. sepium, Cassia tora, Stylosanthes guianensis and Centrosema pubescens were unsuccessful. Volunteer groundnuts growing in open situations were an important source of virus and vector. Field studies on menazon-sprayed and unsprayed plants closely or widely spaced showed that aphid numbers increased rapidly on the groundnut crop 31–45 days from emergence but declined rapidly as the plants matured. Superficially the plants appeared to be free of aphids but large populations were often found on hypanthia, flowers and gynophores. Coccinellids, especially Cheilomenes sulphurea (Ol), were important predators. Aphid numbers were far higher on groundnuts with low plant densities, possibly because the number of favoured feeding sites is increased, but the number of plants infested was not significantly different at high and low densities. Yellow panels had little effect on aphid numbers. Spraying with menazon (294 g active ingredient/ha, 5 times) prevented breeding by incoming alates.

Epidemiological studies of Rhodesian sleeping sickness and its vector Glossina morsitans Westw. were made near Motsowi in the western floodplains of the Okovango Swamps, Botswana, during 1967–68. The low apparent densities of the fly (12–27 in three fly rounds compared with 124 in a fourth located in a more densely forested area to the east) and low recapture rates (1·3–3·9%) indicated the Motsowi area to be a feeding ground. Peak densities in January and troughs in October reflected seasonal movements of the fly to and from their dry-season retreats to the east. Most bloodmeals (61%) were on warthog, but 5·5% were on reedbuck in which the Trypanosoma brucei infection rate was 33%. In the Motsowi area the total trypanosome infection rate in G. morsitans was 5·7%,, with T. brucei 0·6%, but the rate varied between localities elsewhere in the Okovango. Man's movements into the fly areas to the east offer the main pathway for transmission of T. brucei, perhaps aided by seasonal movements of reedbuck closer to the settlements. The incidence of human trypanosomiasis is therefore unlikely to be reduced by insecticide campaigns in the farming areas, but livestock will benefit from a reduction in cattle trypanosomiasis.

The diploid chromosome number of G. austeni Newst. from the Langford colony is 2n = 14 ± 2, + XX/XY. Chromosomes have been karyotyped and include a pair of large submetacentrics (L1), a pair of shorter metacentrics (L2) and a group of small telocentries (S). The X and Y elements constitute a heteromorphic pair. The S chromosomes together with the Y element display allocyclic behaviour during nuclear division. Pre-meiotic RNA synthesis has been followed by autoradiography, and differential staining of DNA with quinacrine dihydrochloride by fluorescence microscopy.

Three new species and one new genus of Eulophidae are described, namely, Wichmannia pictipennis sp.n., reared from small bark beetles living in Euphorbia in the Canary Islands, Myrmokata diparoides gen.et sp.n. which parasitises ants of the genus Crematogaster in Cameroun, and Colpixys gigas sp.n. which destroyed more than 80% of the larvae of a Languriid beetle in the stems of elephant grass in Ghana. The genera Wichmannia Ruschka and Colpixys Waterston previously each included only one species.

Iraqerus neveosparsus sp.n. is described from Quercus spp. in Iraq and a key is given separating the new genus from Stegelytra Mulsant & Rey and Iberia Kirkaldy.

In laboratory experiments with Dysdercus fasciatus Sign, systematic testing of the stylets with a sucrose solution and distilled water has shown that the receptors mediating the salivary response to sucrose are located either on the maxillary stylets or at some site internal to them. Electronmicroscopical examination showed that the maxillary stylets were not innervated, and histological evidence further suggested the epipharyngeal organ as the site of gustatory discrimination. Dendrites located in the mandibular sclerites by means of the electron microscope could possibly have a proprioceptor function.

In laboratory experiments at Langford, Bristol, agar membranes were superior to Parafilm membranes for the in vitro feeding of Glossina morsitans Westw. and G.austeni Newst. Temperature appears to be the primary stimulus for probing but agar membranes stimulate some probing, probably because of the humidity gradient above the membrane. Agar membranes also stimulate gorging, and enhance the effect of the phagostimulant adenosine 5' triphosphate (ATP). Agar membranes may simulate conditions on and in the skin of a living host by stimulating tactile and/or hygroreceptors in the mouthparts of the probing fly. G. morsitans females fed on blood containing ATP concentrations up to 5 × 10−3 M gave a higher reproductive rate and higher pupal weights. The rate of breakdown of ATP in bovine blood during one hour at 37°C indicated the amounts that need to be added to maintain optimal concentrations during feeding. There was positive correlation between pupal weight and reproductive rate of G. morsitans with the membrane-feeding system, but a maximum of only 80% of the optimal rate of reproduction and 80% of the optimal pupal weight was achieved. On living hosts this correlation did not hold, the pupal weights being higher, suggesting that the membrane-feeding system constitutes a stress factor. The stress of in vitro feeding does not apparently involve probing or gorging stimuli but its symptoms are those of nutritional deficiency. It is concluded that post mortem changes in vertebrate blood may affect quantitative aspects of feeding, or they may affect either the nutritional content of the blood or the ability of the insects to metabolise it correctly.

Time/mortality studies with Carabidae exposed to thionazin, phorate, disulfoton and menazon in an acid sandy loam soil were done in controlled laboratory conditions. Carabids were killed by soil treated with thionazin within dosage rates normally required for satisfactory control of nematode or insect pests of crops, i.e., 2·24–8·96 kg/ha, but menazon was virtually non-toxic. Soil moisture increased the speed of kill of Carabidae in thionazin-treated soil presumably by affecting processes of adsorption and desorption of the pesticide to soil colloids and by influencing burrowing behaviour. The speed of kill increased with a rise in temperature, mortalities occurring 2·2 × faster at 10°C, 8·1 × faster at 15°C and 11·8 × faster at 22°C than at 7°C, but was decreased by illumination which decreased the activity of the nocturnal Carabids. The addition of calcium carbonate to increase soil pH did not alter immediate mortality by thionazin but persistence was decreased. Speed of kill was negatively correlated with adult size of Carabid species that behaved similarly, small species such as Bembidion lampros (Hbst.) (mean weight 2·1–2·2 mg) dying 12–13 × faster than female adults of Pterostichus vulgaris (L.) (mean weight 195 mg). Species which burrowed in soil were more susceptible than those which did not, as were newly moulted or starved adults. At 15°C and in wet conditions the “ half-life ” of thionazin based on speed of kill in a soil of pH 6·1–6·7 was 1–2 weeks but when determined by gas-liquid chromatography was 3–4 weeks at 22°C. Leaching could account for the loss of up to 12% of thionazin from the top 9 cm of soil.

Following an evaluation of the anatomical characters and habits of the Scolytini, four genera are recognised, namely Scolytus Geoffroy, Scolytopsis Blandford, Camptocerus Dejean, and Cnemonyx Eichhoff (=Loganius Chapuis, Ceratolepis Chapuis, Coptosomus Schedl, syn. n.). Two South American species, Camptocerus aquilus and C. rectus, having an ambrosial habit, are described as new.

Observations were made on Camptocerus aeneipennis (F.), C. aterrimus Eggers, C. niger (F.) and C. aquilus Wood infesting felled trees in the gallery and valley forests of the north-eastern Mato Grosso between September and December, 1968. The lifecycles of all four species are completed in 7–10 weeks. The gallery systems resemble those of the Xyloterini and Scolytoplatypus. The maternal galleries penetrate the wood and either lie in the transverse plane or run longitudinally, depending on the stem diameter. Trees of the family Burseraceae are especially attacked. The larvae are mycetophagous, developing in cradles which they enlarge as they grow. Larvae were attacked by an ectoparasitic Braconid wasp and predatory beetle larvae. Larval mortality is lower when one or both parents remain in the gallery. The potential forest importance of the four species could be contained by bark removal and rapid drying.

The responses of adult Glossina morsitans Westw. to large slowly moving visual stimuli were tested in the laboratory. The standard stimulus, lasting 60 s, consisted of a vertical black stripe (5° wide) on a white background moved 12 times past a window (25° wide) at 13·5°/s. The kinetic responses (take-off), but not orientation towards the stripe, went through a daily cycle with morning and evening maxima some 500% greater than the mid-day response. Kinetic responsiveness in both sexes and all ages of fly increased in an overall linear manner during five days' starvation, though pregnant females were about half as responsive as mature males. The intensity of orientation also increased with starvation, at least in males. For take-off, the minimum stripe width eliciting responses subtended ca 0·6° to the flies, and the optimum angular velocity was 3–7°/s, for orientation, 2–20°/s; the distance of the stripe appeared unimportant. Human odour elicited take-offs in the absence of the visual stimulus, and enhanced the visual responses if these were tested within three minutes of the start of odour stimulation; these effects were much greater in teneral than in mature males. It is concluded that the differences between the sexes and physiological stages of fly are mainly quantitative, and that the accepted view of four physiologically distinguishable phases of behaviour is therefore unsatisfactory. Instead, it is proposed that tsetse behaviour is the outcome of continuously variable responsiveness to visual host stimuli, and not a succession of behavioural entities.

The response to short photoperiod (10 h) of the F1 and F2 sexual generations of Myzus persicae (Sulz.) resulting from a cross between oviparae from a holocyclic clone and males from an androcyclic clone were examined and compared in experiments done at Imperial College Field Station, England. F1 and F2 progeny segregated into three distinct phenotypes; holocyclic, androcyclic and intermediate. Intermediate clones reared at 10 h photoperiod produced alatae which were physiologically intermediate between gynoparae and virginoparae, because their progeny consisted of oviparae, apterous virginoparae and intermorphs. Intermediacy was not affected by temperature and seems to be genetically determined. Segregation ratios in the F1 and F2 indicate that the androcyclic condition is induced by a recessive ‘ switch-gene ’ which completely suppresses the production of sexual females, and partially suppresses male production. This gene could be present in the heterozygous condition in a large proportion of parthenogenetic lines of the aphid, and thus confer a latent potentiality for anholocycly on populations even where conditions will not regularly permit this method of overwintering. Conversely where anholocycly is strongly favoured many aphids would be homozygous for the recessive, androcyclic character, but the potential for gamic reproduction could remain widespread.

Collections of Anopheles gambiae Giles complex, A. funestus Giles group and other mosquitoes were made by spray-catch from twelve catching stations indoors and by hand-catch from pit shelters at two catching stations at Segera, Tanzania, between January 1970 and June 1971. Females of A. gambiae were identified cytotaxonomically as sibling species A or B of the complex. In houses during 1970, A. gambiae species B was more numerous at first than A, but A became predominant during the long rains of March-May. In the cool dry weather of June-November both A and B densities declined and the A:B ratio surpassed 50:1. The short rains in December produced a population explosion of species B and less multiplication of species A, the B:A ratio reaching >11:1. In 1971, hot dry weather during January-March caused declines of species A and B with a maximum B: A ratio of 13:1. Long rains, coming in late March, provoked a resurgence of A and a concurrent decline of B, so that the A:B ratio again reached 20:1 in June. Similar cycles of species A and B were observed outdoors, although the relative numbers outdoors/indoors averaged 2·3 times more for species B than for species A. In A. funestus, A. gambiae species A and A. gambiae species B Human Blood Indices were 97·5%, 91·2% and 60·9% indoors and 24%, 2% and 7% outdoors, respectively. Respective malaria sporozoite rates were 1·62%, 4·23% and 0·32% and minimum rates of stage-Ill filarial infection were 0·33%, 0·44% and 0·57%. Sporozoite-positive and sporozoite-negative mosquitoes exhibited similar HBF's in species A and discrepant HBI's in species B. The HBI's were higher in filariapositive A and B females than in filaria-negative females. Gregarines occurred in 1·36% of species A and 0·38% of species B. Trematode cysts were seen in two specimens of species A.

Of A and B females 28% and 4%, respectively, had four-banded palps. It is shown mathematically that the discrepant malaria sporozoite rates in species A and B may be explained by extrapolating from the man-biting rate and probable daily survival rate for each species. This implies that no unrecognised factors play a major role in causing the contrasting efficiency of these two sibling species as malaria vectors.

In laboratory experiments individual male Glossina austeni Newst. were each mated, at two- or three-day intervals, to nine different females. In one series of matings the males were restricted to one pairing and in a second series the sexes were left together for 24 h, when more than one copulation could have occurred. The technique followed in the second series was repeated with G. morsitans morsitans Westw. when the males were presented successively with 12 different females. The fecundity of female G. austeni, after mating with males previously mated on 0 to 5 occasions, was consistently higher when left with their mates for 24 h (mean puparia produced per mature female day=0·102) than when restricted to a single copulation (0.085). When the sexes were left together for 24 h, female fecundity of both species was maximal when mated with males used for up to six times (ca 0·10 puparia produced per mature female day) but became progressively poorer as the males were used more frequently (0·022 for G. m. morsitans females mated with males used for the twelfth time). The testes of muchmated males are markedly smaller than those of virgin males of the same age. Maximum disposable output of males from colonies of Glossina engaged in production for the sterile insect release method could be obtained by using males retained in the colony for mating on six occasions. It is speculated that in nature the absence of both teneral and old males from the following swarm ensures the presence of only fully sexually potent males in the swarm.

The effects of soil-applied thionazin on field populations of Carabidae were assessed in two fields of potatoes in Berkshire, England. Broadcast at 11·2 or 44·8 kg/ha in a light sandy loam soil, thionazin considerably reduced numbers of Carabids for up to eight weeks after application. Lower catches occurred again six months later when treated soils were rotavated after lifting of the potato crop; these were attributed to the exposure of Carabidae to thionazin residues which had leached into deeper layers of the soil. Dosages of 2·24 or 8·96 kg/ha applied in rows sometimes increased numbers of Carabidae trapped, possibly because of sublethal effects increasing locomotor activity. Carabidae affected most by the treatments were the small diurnally active species such as Bembidion lampros (Hbst), B. quadrimaculatum (L.), and Trechus quadristriatus (Schr.) which were also abundant at the time when the treatments were applied. Larger species such as Harpalus rufipes (Deg.), Pterostichus vulgaris (L.), P. madidus (F.) and Calathus fuscipes (Goeze) were also affected but were generally less susceptible partly because they appeared later in the season when the pesticide had partly disappeared. The rapid disappearance of thionazin from the surface 1·3 cm of soil could account for the low susceptibilities of non-burrowing species; Asaphidion flavipes (L.) which rarely burrowed was less susceptible and Harpalus aeneus (F.) which readily burrowed was more susceptible than other species of similar size.