High resolution of the Anson drawings: Page1 .

Airspeed had to be about 105 knots to obtain a proper spread and you were supposed to fly at a height of 50 f t in order to obtain a swathe of 70 yards. Nobody ever flew at 50 f t - every pilot knows it is much easier to follow contours at about 20ft! We would fly in loose echelon formation, always on the downwind side to ensure that the smoke was clear of the following aircraft, six aircraft in total, with 100 yds separation between them and at 70 yds spacing span-wise. The sixth aircraft was 200 yds behind the fifth in order to direct position or attitude. When a particular run or area had been covered, the lead aircraft, having come to the edge of the zone, would pull up, do a turn-about and re-formate on the sixth aircraft who was still completing his previous run. The chaps knew the zones perfectly. Nobody had to put any markers or beacons out, each area being named after some local topographical feature, or some such. One shortcoming I would mention was that with the aerosol flowing under gravity. When descending you would get negative "g" and the flow would decrease, whilst the converse was true when pulling up with positive "g". A decent balance could not be obtained. Some areas were under-sprayed and some areas were over-sprayed, and this was one of the factors, which ultimately accelerated the introduction of the helicopter. We had to do the spraying in the early morning, before the thermals arose, otherwise the smoke dissipated too quickly. On occasion we tried evening spraying, but this wasn't too successful due to persistent thermals. We were thus confined to early mornings and the chaps became quite used to getting up in the dark and finishing their job before breakfast time. The heavy breeding areas were in steep, inaccessible, regions, which again contributed to the problem with fixed wing aircraft.

The first phase was carried out in the M'kusi Reserve in the northern part of Zululand, and eight sprays were conducted between August and November 1946, 12 Squadron being employed for this task. It was so successful that they decided to tackle the H'luhluwe and Umfolozi areas, which were seriously affected, the Pongola area of south-eastern Swaziland, and also Gollel, east of Pongola. Later, in 1947, it was extended to Mtubatuba. The success of these operations had to be gauged by some method, and the veterinary means of control was the so-called Harris trap designed by Captain Harris. When the operation started in M'kusi, there were hundreds of these traps, and the average take was about 40 to 50 flies per month per trap. After the eighth application in the reserve, the catch had dropped to 0.2 flies per trap. By mid-1 948 the same situation pertained in the Mtubatuba, H'luhluwe, Umfolozi and Pongola areas, the flies being very difficult to catch, compared with the many thousand which had previously been netted in the Harris traps. So the scientists introduced "bait cattle", which were animals in beautiful condition, accompanied by a few natives, who had to follow along catching the flies. Everything had to be collected, and so in addition to the tsetse you got quite a mixture of things that settle on cattle. The tsetse catch was still higher than that being experienced in the Harris traps and everybody was disappointed.