An epidemic of a dengue-like fever affected the Newala and Masasi Districts of Southern Province, Tanganyika, during 1952 and 1953.

The topography, climate, water supply, vegetation and population of the area affected are described.

The disease was most active on the Makonde Plateau in Newala District but the surrounding lowlands were also affected. The plateau, lying at about 2,000 ft. above sea level, is of sandstone, highly porous and practically devoid of surface water. The plateau population, nevertheless, is dense and the people habitually store water in their houses which in consequence are infested with the mosquitoes Aedes aegypti and Culex fatigans.

The first cases of the epidemic were in Mchichira on the south-eastern fringe of the plateau in July, 1952. The densely populated south-western part of the plateau was not involved until about the beginning of September, but thereafter spread was rapid. The peak of the epidemic was in January, 1953, when 49 of 62 localities were involved simultaneously. The decline of the epidemic was rapid after this time and by mid-March only 22 of these localities reported cases of the disease.

Among 13 plateau localities the crude incidences of the disease varied between 13 and 95 per cent. Similarly, among eight lowland localities, incidences varied between 0 and 37 per cent. In plateau localities in contact with woodland, the incidence in adults or adult males was significantly lower than in children. Escape from attack was most common among males over 45 years of age, and was not associated with travel to coastal localities. As there is no reason to believe that adult males were less likely to acquire infection in the epidemic, this is taken to indicate that these males were protected against the disease by some previous local experience not shared by the women and children.

The incidence of the disease was related to houses: once infection had been introduced into a house then all the inmates tended to be attacjed. The initial subject was either an adult or a child.

The results of catches of insects in houses are reported and the methods by which laboratory animals were inoculated in attempts to isolate virus. Virus was isolated from all the mosquito groupings used (Anopheles spp., A. aegypti and C. fatigans), and possibly also from bed bugs (Cimex hemiptera). Full details of these virus isolations will be presented in a letter paper (Ross, in preparation). The occurrence of the various arthoropods in the houses and their habits are studied in relation to the disease incidence; the evidence points to A. aegypti as the vector of the disease.

Some other related entomological work is reported and a complete list of the biting arthropods encountered is given.

The genesis and spread of the epidemic is discussed. It is considered likely that the disease originated in the Ruvuma Valley and was spread about the plateau by natives visiting between hut groups. The behaviour of A. aegypti is discussed in relation to virus spread; its habit of biting on the verandah of huts by day is considered important.