From 1940 up to the time of writing, experiments have been carried out in the savanna country in the north of the Gold Coast on the trapping of Glossina palpalis and G. tachinoides, which are the most important vectors of sleeping sickness in West Africa.

The habitat of these tsetse is described, with special reference to their feeding grounds, since those are the places where, trapping can be most effective and control of the flies is most desirable.

A new type of trap, the Animal trap, was designed to meet the feeding habits of the two species of tsetse whose main hosts in the country where sleeping sickness is prevalent are man and domestic animals.

The construction and maintenance of the Animal trap are described, the essential feature of which is a barrel-shaped body standing out well under good lighting and easily visible from all angles.

Traps with a standardised 2-feet-long body raised 1 foot from the ground combined optimum catching with convenience in construction and portability.

Light brown or khaki covering was slightly better than black, and rough material was decidedly better than smooth. Under poor lighting conditions these differences were less marked.

Comparisons over long periods showed that the Animal trap caught over 20 times as many flies as the Harris trap which was not at all suitable for G. palpalis; about 10 times as many flies as Swynnerton's A.S.B. trap; and about seven times as many flies as Chorley's Crinoline trap.

The correct siting of traps is of the greatest importance.

Traps catch well anywhere in fly-belt but are most effective in open feeding grounds such as ferries, fords, waterholes and animals' drinking places.

The regular presence of an abundance of hosts is the, most important single factor for good siting, and the nearer the traps can be placed to natural hosts the more effective they will prove.

Traps should be placed if possible with their axes at right angles to the regular lines of movement of the flies.

Good visibility of the trap from as wide an angle as possible is of great importance. Traps about five yards from the edge of the fly-belt catch twice as well as traps right at the margin of the bushes and considerably better than traps 40-50 yards out.

Good lighting, especially from overhead, is an advantage but by no means essential.

Good sites remain consistently good as long as the environment is unchanged, and bad sites remain bad.

There is a tendency for the numbers of flies caught to fall by as much as 70 per cent, during the first few weeks when traps are placed in undisturbed fly-belt. This is the effect of skimming off a part of the fly population in the immediate vicinity of the traps and does not represent a real reduction in population.

Comparison of tsetse catches by groups of five traps operating continuously, and by flyboys at the rate of 20 flyboy-days a month from parallel series of observations, gave the following facts for which explanations are offered.

Traps relative to flyboys are more efficient in taking G. tachinoides and less efficient in the cases of G. palpalis and G. morsitans. Flyboys, however, catch a higher proportion of female G. tachinoides but a lower proportion of females of the other two species.

It is considered that Animal traps represent the regular hosts of G. tachinoides and so attract non-feeding males as well as hungry flies of both sexes, whereas in the cases of G. palpalis and G. morsitans the traps represent something outside the range of their preferred hosts arid so attract only hungry flies bent on obtaining a meal.

Traps appear to have a greater efficiency relative to flyboys in groves and clearings where the numbers of flies are small, than in sites in or near heavy and extensive fly-belt with high fly populations. This is possibly the result of competition which would be most evident in small tsetse populations and which would act in favour of the continuously present traps.

There is a striking difference in the annual rhythms of catches by flyboys and by traps; the former showing a double-peaked curve with maxima at the beginning and in the middle of the wet season, the latter with a single peak towards the end of the dry season and a minimum during the rains.

The efficiency of traps relative to flyboys is greatest in January and least in September. In an estimate based on all the available data and comparing 150 trap-days with 20 flyboy-days monthly, the traps at their best were 10 times as good as the flyboys, and at their other extreme were as low as one-fourth of the flyboys' catches.

It is considered that this maximum efficiency in the dry season is not due to the flies' attraction to shade but that it is a complex result of the circumstances which bring about optimum conditions for trapping in the dry season, when the weather frequently puts flyboys at a disadvantage, and the most adverse conditions for trapping in the wet season when they enjoy certain advantages of mobility.

The seasonal rhythm of trap catches has demonstrated the presence of a considerably greater fly population in the height of the dry season than is suggested by the magnitude of boys' catches.

Three possibilities of trapping in the control of sleeping sickness are examined : (1) for extermination of the tsetse; (2) for reduction of the fly population; (3) for destruction of feeding flies in contact with people.

Regarding the first possibility it had been found that continuous catching by teams of four flyboys could eliminate G. palpalis and G. tachinoides from small groves after 8–12 weeks' continuous catching but that, since none of these groves was completely isolated, repopulation took place by immigration, especially in the wet season. Small numbers of traps, 5–7 to the acre, made no apparent difference to limited tsetse populations but concentrations of traps up to 20 to the acre brought about considerable reductions and checked the effects of immigration, but with no approach to extermination of the flies.

Eradication of the tsetse by traps is impossible but reduction of fly populations, if it can be effected over the main area of an epidemic, should, in theory, result in a partial control of trypanosomiasis.

Small groups of traps well sited in feeding grounds bring about considerable reductions in the numbers of active, hungry flies in contact with their hosts although hardly affecting the main fly community. This, in theory, will have a definite protective value to the community, enhanced by the fact that traps show increased catching powers when in close proximity to man and animals and that there is a tendency for people and tsetse to concentrate around waterholes (thus coming into the most intimate contact), during the dry season, the very time when traps are working at maximum efficiency.

It is in this protective role that traps seem to show their greatest possibilities for use in the control of sleeping sickness. The measure is applicable anywhere. It needs only small numbers of traps at each place where infection is being transmitted so that a reasonably big area could be covered economically. It might be a useful adjunct to clearings which fail to exclude the tsetse sufficiently. Traps so used might prove a valuable supplementary control-medium in places, such as the forest, where clearing is unlikely to be easy or cheap. Finally, experiment has shown that the retention of toxicity to Glossina of DDT-impregnated hessian exposed to the weather is sufficient to justify the application of DDT to the material covering the traps in order to increase their killing power.