Living specimens of Megarhinus brevipalpis were transported from southern Natal to Johannesburg to establish an insectary-bred colony. The natural habitat of these predatory mosquitos consisted of small isolated patches of sub-tropical forest, in which the rainfall is 40–50 ins. (102–127 cm.) with a mean winter temperature of 64°F. (17·7°C.) and an annual range of 27°–33°F. (15°–18°C). The breeding places were leaf axils of Strelitzia nicolai (a plant resembling a wild banana), small rot holes in trees and larger ones in Strelitzia stumps. The larvae were collected from leaf axils with an apparatus consisting of a rubber bulb to which were attached lengths of glass and rubber tubing.

The insectary was a room 9 ft.×8 ft. 6 ins. and 9 ft. high which was kept at tropical heat and humidity. Mating of the adults was observed, copulation being effected while at rest or in flight. Oviposition was usually accomplished in flight but also while at rest on the surface of the water. In the summer time two females, which were tested, laid about 85 eggs each during the month following emergence from the pupa, six or seven days elapsing after emergence before the first oviposition. In the middle of the winter, oviposition (with later generations) became very irregular in spite of the temperature and humidity remaining constant. The adults, which were comparable to those of the natural habitat, were fed on sugar solution, honey and fruit juice. One bred out as a gynandromorph.

When given an abundant supply of larvae of laboratory bred Aëdes aegypti, the life-cycle of M. brevipalpis was normally : egg (incubation), less than two days ; larva, 11–20 days (average 14·5 days) ; pupa, five days. This does not include a small number of exceptional cases in which the life as a fully grown larva was abnormally prolonged (in one case nearly four months) for reasons which are not absolutely clear. The larvae killed from 100 to 200 or more Aëdes larvae during the normal larval life, but many of these were not eaten when the brevipalpis were in the late fourth instar. By a special technique they were also induced to eat dead tissues including minced pork brawn, minced maggots and minced flies. Except for the latter these were not satisfactory foods although there was slow development.

Fourth-instar larvae were kept out of water for three to four weeks (without food), in a damp atmosphere, and afterwards when fed most of them developed normally, but pupation was sometimes suspended for a considerable time. They have been sent by post (out of water) in tubes with damp cotton wool and filter paper.

The egg differed from that of other Megarhinus species in having a crown of projections at one end with a cup-like structure in the centre. The exochorion had roughly hexagonal cells but without numerous tubercles as in other species.

First-instar larvae remained in the egg-shell after hatching when the eggs-were out of water but on a damp surface and in a saturated atmosphere. They survived like this for up to six days or about the same time as the larvae survived in tap water if there was no food. When liberated in water the head of the first-instar larva was comparatively small with the mouth parts folded in. Within two hours of liberation in water the head enlarged considerably and the mouth parts came into position ; the larva was then ready to catch its Culicine prey. When in water containing dead leaves, these larvae survived from a few days to over four weeks and some grew to the third instar without any Culicine food.

Cannibalism was investigated. Fourth-instar larvae did not attack each other readily ; they devoured smaller larvae of their own species and small to medium size larvae resorted to cannibalism, particularly in the absence of Culicine prey. There was evidence that fourth-instar Aëdes aegypti occasionally ate first-instar Megarhinus.

The discussion traces attempts which have been made in certain Pacific islands, notably Hawaii and Fiji, to use Megarhines for biological control of disease-carrying mosquitos. M. brevipalpis has a shorter life-cycle than the species introduced into these islands and the conclusion reached is that laboratory breeding, to enable large numbers to be released in certain areas, would be a suitable adjunct to a programme of general control, in this part of the world. Airmail consignments of larvae are being sent to Hawaii with the object of starting a laboratory colony there.