During July and part of August 1929 an experiment was carried out at Sambala (Western Kondoa), Tanganyika Territory, designed to show that : (i) Hungrier flies marked in Berlinia- Brachystegia woodland left that type for the supposed feeding-grounds (“ female centres ”) earlier than repleter flies ; (ii) conversely, repleter flies remained within the woodland longer than hungrier flies ; (iii) hungrier flies in the supposed feeding-grounds remained in them for the first few days ; and lastly (iv) repleter flies in these situations returned to the Berlinia-Brachystegia woodland to remain there until later.