THE TSETSE PROBLEM IN NORTH MOSSURISE. 359 



That homing occurs, within limits, is to me certain. On 29th June as soon as 

 J (^ntered a male crowd the flies as usual attached themselves to me and kept pace 

 with me, buzzing about my feet. On my leaving the vlei and entering the Brachy- 

 stegia their numbers fell ofi rapidly, only a few following me 50 yards in. I stood 

 still finally when all had left, and after a few minutes returned to the cluster to see 

 if I should pick up the flies along the path, where they had left me. I saw not one 

 fly till I had got to the cluster, in which they seemed as numerous as before. It 

 seemed probable that on leaving me they had flown straight back to the cluster. 

 Natives arriving with loads shortly afterwards passed through the cluster and a 

 number of the flies accompanied them right through to my camp — a further 150 

 vards — possibly a matter of preference. They stayed for a short time, occasionally 

 biting, then all disappeared. 



Similarly whenever we emerged from a cluster, followers would just at first often 

 be numerous, but for the most part would fall off within a very few hundred yards, 

 and on the analogy of the above observation and from certain takings of marked 

 flies, it seemed likely that these short-distance flies had flown back. The numbers 

 of the remainder lessened rapidly with the distance travelled and relatively few 

 followed for some miles. Would these long-distance flies also find their way back 

 unaided ? Would they necessarily proceed further than the first useful vlei ? And if 

 deposited in congenial surroundings, would they necessarily return at all? The 

 experiments known to me on the subject of long-distance homing do not (as described) 

 seem to settle conclusively the question as to whether it was unaided. Simpson's 

 were on highways of traffic. His recaptures were many and confined to the vicinity 

 of these highways — road and river. Lamborn's recaptures were few and his general 

 distances longer. From my cattle experience I judge it to be likely that fly will go 

 farther on game than on man. They also, in my observations, tended to travel farther 

 through unattractive country, once there, than through country offering temptations 

 to dismount. This suggests further the need for a closer description of the country 

 searched on the wrong side of the point of release. The question is of some possible 

 practical importance, and it was, therefore, a matter of regret to me that lack of 

 time, with the fact that on first leaving the Kanyezi centre I did not know if I could 

 return to it, prevented me from carrying out long-distance experiments of my own. 



My incidental evidence suggested that homing does not always take place at once. 

 The fly that I have described as travelling six miles in three stages reattached itself 

 each time at the spot at which it had left one or other of us the day before, and 

 other isolated instances of the same thing occurred. Of my marked flies that were 

 carried to native villages a few miles of[ two were reported as staying about for from 

 two to four days before disappearing — when they may merely have moved farther 

 on on natives or have been killed. Further evidence was obtained in investigating 

 a severe and sudden outbreak of nagana in a large herd in British territory, verv 

 many miles from fly (permanent or otherwise), which appears to have been caused 

 by male tsetses carried by game. It had taken place at the commencement of the 

 dry season. The native in charge and the herd boy, interviewed separately, each 

 picked out tsetses (from amongst my specimens) as having been on the cattle before 

 and during the outbreak and stated that they had appeared suddenly and stayed three 

 weeks or rather more. They disappeared " when the leaves fell " — a spontaneous 



