An incomplete ring of electrified nets was placed round a trap or round an electrified net and visual target in woodland infested with Glossina morsitans morsitans Westw. and G. pallidipes Aust. in Rhodesia. The distribution of catches in these systems was used to estimate the minimum efficiencies with which the trap or net and target captured flies that were initially attracted by odours derived from a herd of cattle hidden below ground. With mature tsetse, the estimates for the trap were 23 and 21% for male and female G. morsitans, respectively, and 70 and 49% for male and female G. pallidipes. For the net and target, the corresponding figures were 29, 30, 53 and 43. The coefficient of variation of such estimates (about 10%) was low enough to suggest that the technique affords a quick and reliable means of screening the efficiency of many trap designs. Factors responsible for the efficiency of the trap were elucidated. Data for other insects were obtained.

The identity of Empoasca dolichi Paoli, a serious pest of beans and other crops in Africa, was confused. While processing a collection of leafhoppers from Malawi, E. dolichi was found mixed with other species. The identity of E. dolichi was established with reference to its type material, and two new species and a new subspecies of Empoasca from Malawi and two new species from Nigeria are described. New characters that might help specific identification and show the relationships of groups of species are also illustrated and described.

The causes of fluctuation and stabilisation in a population of Andraca bipunctata Wlk. on tea in India were assessed over three years using the technique of key-factor analysis. Failure to produce the maximum complement of eggs, together with female mortality and adult migration, is shown to be the key-factor. Other mortality factors include non-viability of eggs and parasitism of the caterpillars by the tachinid Cylindromyia sp. Mortality of the female pupae tends to stabilise the population by compensating for losses due to the key-factor, but neither this nor any other mortality factors appear to be significantly density-dependent. The need for devising specialised sampling procedures and statistical tests for tropical insect populations is discussed.

Hybrids of Nephotettix virescens (Dist.) and N. nigropictus (Stål) have been bred in the laboratory, and they could transmit rice tungro virus even though only one of the parents is a vector. In the present paper, evidence is presented based on coloration and on male genitalia that suggests that such hybrids occur in nature. The potential danger from such hybrid vectors is pointed out, and taxonomists are alerted to the risk of accepting such hybrids as new species, thus completely hiding their potential as vectors.

The Polovodova ageing technique was applied to females of Anopheles darlingi Root in the northern part of Mate Grosso, Brazil, in April 1978; 1·5% of the females had oviposited at least four times and were old enough to be potential vectors of malaria. The oldest female dissected had oviposited six times. The mosquito has a resting period of at least 24 h between oviposition and the next blood-meal. There was a pronounced peak of biting activity, consisting largely of nulliparous females, at dawn and dusk.