1. The female tsetse extrudes her larva on soils ranging from fairly heavy clay to coarse sand; the presence or absence of organic matter seems to be immaterial.

2. The soil temperature varies greatly in different breeding-grounds in the dry season, but little in the rains.

3. The soil water content varies greatly in different breeding-grounds in the rains, but little in the dry season.

4. The soil water content in some of the breeding-grounds falls so low in the dry season that there can be no doubt that the atmosphere in these soils is below saturation.

5. Evaporation measured at 5 inches above ground-level varies little from site to site; bigger and important differences occur at greater heights above the ground where the screening effects of thicket become operative.

6. The seasonal changes are so great at Gadau that no one breeding-ground can satisfy the female's requirements throughout the year; instead, different breeding-places are selected for different seasons.

7. The cycle for G. morsitans is as follows: In the rains breeding occurs under palm fronds and logs in the open woodland. In the early dry season breeding continues in the open woodland, but shifts from the palm and log sites to the small thickets; breeding also starts in the more open parts of the forest islands. At the beginning of the very hot weather all the thickets of the open woodland are evacuated, and breeding is confined to the densest parts of the forest islands. In the early rains the movement is reversed: breeding shifts out into the open woodland, and in the heavy rains is confined to the log and palm sites.

This cycle fits in closely with the cycle for the seasonal concentration and dispersal of the adult population (Nash 1937, pp. 85–90).

8. The major wet season breeding-ground of G. tachinoides is unknown, but the cycle is believed to be as follows: In the early dry season breeding shifts from the log, palm, and probably from the unknown site, to the small thickets and forest islands, and becomes maximal in the cold season. Breeding now decreases in all sites, suggesting that none is really suitable—a surmise which is strengthened when tremendous breeding activity starts in the river-bed, as soon as it becomes available at the beginning of the hot season. In the early rains this site is destroyed and breeding is believed to commence in the unknown site, as it certainly does under the logs and palms.