374 C. F. M. SWYNNERTON. 



G. pallidipes and Tabanus had never before been noted on them in tliis dense 

 wooding, though they occurred in the bush outside. Yet in these three hours nine 

 pallidipes (all but one females) and 31 Tobanus were taken on them. 



I then turned the two oxen just into the uncleared undergrowth alongside and 

 they were at once attacked by brevipalpis. My collector captured 15 on one of them 

 here in fifteen minutes and no pallidipes or Tabanus were seen at all. 



The effect of clearing the undergrowth alone in banishing the big fly was, I think, 

 well demonstrated. It is perhaps more likely that the other flies were attracted 

 to the cattle from the bush outside than that they had taken up their abode in the 

 forest as the result of the clearing of the undergrowth. This would suggest that 

 attacks of pallidipes in the sunny hours are undertaken from a greater distance 

 than those of brevipalpis, being equivalent rather to the latter fly's evening attacks. 



Experiment 2. To ascertain the width of undergrowth clearing necessary to protect 

 a strip of road from attacks by brevipalpis in sunny weather. In Brachystegia bush on 

 the Buzi, the cattle in passing along a certain track four times in all had on each 

 occasion been attacked by brevipalpis all the way along from the thickets passed 

 closely or passed through. I cleared the dangerous undergrowth, but none of the 

 overwood, for a paced length of one mile and a width of 25 to 40 yards with the 

 track in the middle. I left three small points not fully cleared — some small Diplo- 

 rhynchis trees with low^-hanging branches and Bauhinia, a small prmiary-type 

 thicket, and a piece of uncleared sapling growth. I put the cattle through and they 

 were attacked at the first two of these places — nowhere else at all. I brought them 

 back just outside the clearing and they got flies all the way. In the afternoon, 

 having cleared these spots, I put them through agam, bringing them back this time 

 along the track. They got no flies at all either way until driven out beyond the 

 cleared strip at each end. They were then each time attacked at once. 



It was to be expected that this narrow clearing would be of use only in sunny 

 weather and this point was tested later on an overcast day. The cattle w^ere driven 

 along the cleared mile and 32 brevipalpis were taken at them, coming in all the way 

 along. They still failed to be attacked here in sunny weather. 



Obviously then, very little clearing is needed to safeguard a given piece of road 

 from this fly, provided that the cattle pass along it in sunshine. A considerably 

 wider clearing would be needed to render it safe at aU hours and in a^l weathers. 



Clearing by Arsenite of Soda. I had previously carried out successful experiments 

 in poisoning with arsenite epiph}i;ic figs in forest on my own property. This sug- 

 gested its use in clearmg for tsetse, for the difficulty in ordmary axe-clearing is 

 that the stumps at once (and repeatedly) send up fresh growths and form coppice 

 that is even more suitable for the fly than the growth cleared. A recent examination 

 of some of my experiments shows that the strength of an ounce of arsenite to a 

 pint of water (the maximum strength tried) is insufficient to be generally effective, 

 though it killed some trees and inflicted damage on many others that may enable 

 the fires to burn them out. I hope shortly to be in a position to state the best 

 strength for this sort of work. Given this and the amount of liquid required per 

 acre for particular kinds of bush, one can estimate very roughly the cost of this 

 kind of clearing. 



