THE TSETSE PROBLEM IN NORTH MOSSURISE. 383 



the margins of the " Oblong," and it is unfortunate that such fires should necessarily 

 have ceased at the very moment when whole districts of native gardens were 

 abandoned to grow up into coppice growth. 



The effect on reproduction of burning at a time when a large proportion of the 

 trees are in flower must also not be overlooked. I wrote on 5th December 1906 

 on examining a late " burn " between Chibabava and Arucate and observing 

 that the flower crop generally had been destroyed, "it will be interesting to note to 

 what extent this late burning will effect the seeding, and consequent replacement 

 of dead individuals." There can be little doubt that it will affect it considerably 

 in the lower types of wooding. 



Late burning is no emergency measure against fly. In some areas especially, 

 where the grass is already well reduced, its full effects may take long to show. Again, 

 its abandonment will mean the return of the old conditions, for the underground 

 stumps and roots of pyrophytes are extraordinarily tenacious of life. It must be 

 regarded as a piece of administrative policy and be kept up annually. That its 

 adoption is necessary I am quite convinced, if only for the reason that the country 

 generally is at present reverting more and more to wooding and becoming increasingly 

 fitted for tsetse, even in those areas in which the cover has hitherto been sparse and 

 light. The recommendation of the measure, in face of its having to be kept 

 up indefinitely, must be that the fires take place in any case and that the annual 

 cost of regulating them will be trivial. The latter may even be met by the fines 

 imposed on unauthorised burners, and the actual burning, being done at a signal 

 by the kraal natives themselves, will cost nothing. 



Apart from the prophylactic value of the measure just mentioned, its effect, grass 

 conditions being equal (which they are not), should show first and chiefly in relation 

 to G. brevipalpis and (in the area in which this fly depends mainly on coppice) 

 pallidipes. It will not affect the grown pyrophytic trees in any wholesale manner 

 (unless in a very exceptional season or unless they are weakened by ring-barking). 

 Its effect in relation to morsitans might even be regarded as problematical, for that 

 fly and pallidipes in the same wooding appear to be independent of undergrowth. 

 The systematic burning back of smallish growth will assuredly tell as the established 

 trees pass maturity and eventually go, with nothing to replace them, but the time 

 involved will be long. 



The objections that may occur should be discussed. One is that the loss of the 

 wooding at springs may dry these up. Many of the springs rise already in open 

 vleis, and wooding on damp ground (including big Khaya and Adina) is in any case 

 not easy to destroy by burning, except occasionally in such a season as 1913. But 

 better an occasional spring lost than a continuance of the tsetse ; and the Zulu 

 •clearings, sufficient to remove the fly, do not seem to have caused any shortage of 

 water. At worst, occasional lost spring-heads could be allowed to revert to wooding 

 -after the tsetse had been eliminated. The exposure of the ground to wash by burn- 

 ing shortly before the rains is not an effective objection, as it has been our experience 

 here that the growth of spring-burned grass rapidly overtakes that of winter-burned 

 grass, which itself affords little protection to the soil in October. Loss of, or damage 

 to, the secondary wooding generally — unless of the very finest Albizzia-Milletia- 

 Pterocarpus monsoon types — ^need not be regretted, for reasons which I have stated 



