In eastern Africa between 1951 and 1955 and in south-western Morocco in 1960, orientations and densities of flying locusts in migrating swarms of Schistocerca gregaria Forsk. were recorded photographically, with supplementary visual observations. The densities ranged from 0·0006 to 1·23 locusts/m3 (mean spacing between nearest locusts 12 and 1 m, respectively) with a median value of 0·024 locusts/m3; in 50% of the samples mean spacing was 2–4 m, suggesting a preferred distance of that order. The orientations were affected by gregarious reactions to neighbouring fliers, and to locusts in a mass. In the swarms there was a preponderance of streaming groups in which locusts headed in the same general direction; in about half of these groups the angular deviation about the mean orientation was only ±20°. In comparatively dense swarms the orientations of the lower fliers in the interiors of swarms were diverse and random in relation to wind and to the swarm track. But at the leading, the trailing and the lateral edges they were predominantly into the swarm, as the result of observed changes of headings back towards the swarm in locusts flying out beyond the main mass. The orientations of locusts photographed at about 100 m and 200–400 m above the ground were found to be predominantly in the direction of the swarm track, i.e., with wind; it is suggested that they arose from compensatory optomotor responses to retinal velocities of substrate patterns, and are consistent with Kennedy's hypothesis of optomotor orientation to wind. Among lower-flying locusts such responses appeared to become important in diffuse formations and in strong winds. Thus flight on consistently into-wind courses was repeatedly seen in low-flying diffuse formations of sexually mature locusts migrating in light winds. In stronger winds the orientations of lower fliers in all types of swarms suggested responses to unacceptable retinal velocities, which interfered with the into-the-swarm orientations on the swarm edges; the continued cohesion of swarms may thus require a wind profile in which they are not evoked below a certain level. In persistent strong wind in excess of locusts' air speeds, flight activity was reduced, and swarms exhibited low into-wind flight by locusts in small numbers. It is suggested that the characteristic downwind tracks of discrete immature swarms are set primarily by the downwind orientations of their higherflying members. In sexually mature swarms the downwind displacements are probably performed mainly by the denser and the higher-rising swarms, while the into-wind flights of low diffuse formations would slow down movement, and could result in limited displacement up-wind.

The Varley & Gradwell method of population analysis is applied to data on populations of Saccharosydne saccharivora (Westw.), comprising 51 generations from 21 fields, obtained in Jamaica in 1961–62. Populations were estimated at five points in each generation, namely, maximum potential natality, and numbers of eggs laid, instar 1, peak instars 3–5, and of normal reproducing adults. Taking maximum potential natality as the first point in each life budget, graphical analysis showed that the key factor was mortality between instar 1 and reproducing adults or between peak instars 3–5 and reproducing adults. Reduced fecundity, due to submaximal leaf nitrogen, was density independent or possibly inversely density dependent; one egg parasite, Tetrastichus sp., was inversely density dependent while another, Anagrus flaveolus Waterh., and an undetermined egg mortality factor were density independent; parasitism of nymphs and presumably adults also by Stenocranophilus quadratus Pierce was density dependent, and was Considered to be the only regulating factor. The key factor was density independent, and it is concluded that, of the separate mortality factors contained within it, dispersal of adults is the most important. This itself depends on crop characters which could be altered by selective breeding or agricultural practice.

A comparison of the mosquito fauna in a forest near Zika forest, Uganda, in July-August 1971, showed that Mansonia (Coquillettidia) fuscopennata (Theo.), Culex (C.) annulioris Theo. and M. (Mansonioides) africana (Theo.) made up 84% of the catch taken at 24-h human baits in a horizontal transect, whereas M. (C.) metallica (Theo.) made up 90% of those taken at rest. Precipitin tests showed that 98% of M. metallica and most individuals of the other Coquillettidia spp. had taken avian blood, but that two out of seven individuals of M. fuscopennata had taken primate blood. In olfactometer studies M. metallica did not respond to the human hand or a chicken. It is concluded that none of the Coquillettidia spp. caught resting, except M. fuscopennata, has any preference for man.

The effect on the size of sample of 32 taxa of Macrolepidoptera, and on the index of diversity α, of various modifications of the light in Rothamsted and Muguga light-traps was investigated in experiments at Muguga, Kenya, in 1968. The sample of moths from the standard Muguga mercury-vapour trap was fifty times as big as that from the standard Rothamsted Tungsten trap. Using a mixed mercury-vapour and tungsten lamp in a Muguga trap halved the sample, as also would obscuring 60% of the light from a standard mercury-vapour lamp. Overall size of sample, direction of illumination and source of illumination all affected the proportion of different taxa in the sample, but the diversity index differed little in any of the combinations of traps, lamps and sites. Analysis of the pattern of sample distribution over the site showed that moths of all sizes were migrating downwind and height of flight was found to be positively correlated with size. Interaction between traps and sites, combined with the differing heights of flight of the different taxa, complicated the definition of specific sample size; a three-dimensional representational model attempts to define the spatial population parameters required to define the specific sample.

The distribution of the larvae and pupae of Culicoides melleus (Coq.) in a natural beach habitat in an inlet on the Florida coast was studied in relation to tidal level and time of day. Larvae and pupae were confined to the intertidal zone and these distributions were unaffected by time of day or tidal level. Distributions of younger larval instars were displaced towards the inferred area of oviposition at higher elevations on the beach. Pupae were associated with the zone at and somewhat below high tide level. Differences in the vertical distribution of larvae between 0 and 2 in. deep between dark (pre-dawn)/low tide, morning/low tide and afternoon/low tide conditions were probably associated with larval response to light and heat.

In laboratory experiments with a temperature gradient, second- to fourth-instar larvae preferred the 18–25°C range, but the fourth instars were the least discriminating. Laboratory experiments with a simulated beach showed that pupae flooded on an incoming tide remained in their burrows and immediately buried themselves more deeply when waves were generated in the water; they were able to return to the surface if buried, and survived drowning for four days. Inundated areas were avoided as pupation sites. The orientation of pupae at the sand surface and their formation in burrows was related to negative phototactic responses of the pupating larvae.

On the basis of specimens from New Guinea and New Britain, the genus Dynamenais Kirkaldy (Pyrrhocoridae) is re-defined, four species and one subspecies are described in the new subgenus Syncrotus (Syncrotellus) (Pyrrhocoridae), and two species are described for the new genus Goilalaka (Pentatomidae). S. (S.) similis sp.n. was recorded from cacao, and S. (S.) madanganus sp.n. and G. rubroruga sp.n. from Solanum spp.

Observations on the behaviour of the predatory Reduviids Phonoctonus fasciatus (P. de B.) and P. subimpictus Stål were made in the laboratory. The sequence of events in the response to normal prey (Dysdercus superstitiosus (F.)) is described. The preferred site of the first puncture with adult prey was an antennal base, or the same region and the neck with nymphs. A strong tendency for the Reduviid to move its proboscis tip towards the prey's head region implied some perception of body form, which was found to depend on the presence of the prey's appendages. Such perception had no apparent influence on the intensity of the predatory response, experiments with models showing the most important visual stimulus eliciting attack to be that of movement. The scent of the prey greatly augmented the response to models, and chemical stimuli are presumably the basis on which the prey is selected and conspecific individuals, which resemble the prey in appearance, are recognised. The rarity of cannibalism by the adults showed such recognition to be well established, but sexual discrimination was minimal, both males and females behaving almost identically towards either sex. Adults of the two Phonoctonus species also failed to discriminate between their own and the alternative species, but since attempts at cross mating were unsuccessful their separate specific status seems justified.