In order to evaluate Morris traps as a method of investigating populations of Glossina pallidipes Aust., and to compare them with fly-rounds, a series of catches was made by both methods in and adjacent to an area of evergreen thicket covering about three sq. miles in the Lambwe Valley, South Nyanza District, Kenya. During the period of the investigations (February 1955-February 1958), the mean monthly apparent density (catches of non-teneral males per 10,000 yd. traversed) determined from a fly-round 5,200 yd. long traversing the thicket varied from 70 to 1,300, and the mean monthly trap catch, determined from 26 traps along a similar traverse but over a shorter period, varied from 190 to 670. Catches by the two methods showed similar fluctuations but were not significantly correlated, and it is not certain that real changes in population occurred. The distribution of G. pallidipes between the three main vegetation communities sampled appeared different when studied by traps, or by other methods (fly-rounds, and searching for resting flies). The two latter methods were in agreement, but traps gave different results, perhaps because one of the vegetation communities may have had a high proportion of productive trapping sites.

Dispersal of G. pallidipes into the surrounding, sparsely wooded grassland was studied by traps arranged in lines extending east and west of the thicket and at right angles to its edge. Traps appeared not to attract flies over a distance as great as 100 yd., but to catch only those that chanced to be in their immediate vicinity. Total catches in three traps over 10 months at distances of 5–100–300–500 and 900 yd. west of the thicket edge were 15,417–965–153–89 and 56, respectively; smaller numbers were taken up to 2,500 yd., the greatest distance investigated. Dispersal was greatest in the wet, cool half of the year.

In catches of G. pallidipes made in six consecutive two-hour periods daily for one year in two batteries, each of seven traps, and on two short fly-rounds, one of each along the east and the other along the west edge of the thicket, the traps yielded 78 per cent, females and the fly-rounds only 17 per cent. There was a marked tendency for females to be caught earlier in the day than males by either method, and for the pattern of catches to be earlier on fly-rounds than in traps. No difference in age, estimated by wing fray, was found between flies caught, by either method, at different times of day.

The total number of G. pallidipes caught by two traps operating 11 hours per day for 10 days at the thicket edge was over four times that caught there concurrently by a stationary party of three men.

It is concluded that traps are valuable because they catch a high proportion of females, thus affording information not given by fly-rounds, and operate continuously, at a higher over-all catching rate than that of men, thus facilitating the study of sparse populations. Nevertheless, site effects and day-to-day variability are both large with traps, so that reproducible results are difficult to obtain.