go to previous section of this publication
The general scheme was for the scientific party to arrive at the swarm roosts before sunrise, and to start observations at once. Air temperature would then be fairly close to the night minimum. Frequent records of the major variables would be made. up to the time when all the locusts left the site. One man was responsible for air temperature and humidity (sling and meteorological screen), and wind speed and direction, and another for 'radiation; one man was continuously employed taking locust internal temperatures and another caught locusts for him; the fifth was responsible for the observations on locust behaviour. Illness or camp duties usually accounted for a sixth member of the party; on one occasion; only two. were fit for work.
As far as possible, the instruments were placed in a good position exposed to the rising sun and in a good concentration of locusts; but it was sometimes necessary to give the meteorological instruments an eastern exposure, where there were no locusts, and to place the thermocouple, which was used for measuring locust body temperature, some distance away and amongst the nearest dense locusts. It cannot be claimed that the positions chosen were representative for the whole swarm.
As soon as all the locusts had gone, most of the party would return to camp, but relays of light vehicles would keep in touch with the swarm until it roosted again for the night. If the swarm was lost, several days might be spent in finding another, but if contact was successfully maintained, the same swarm would be studied again next morning. Between 24th April and 11th May, one large swarm was followed from 00°40'S., 35°11'E. to 02°00'N., 35°27'E. (Fig. 1) and detailed observations made on it on 10 mornings; such success was unusual and would not be possible with a small swarm or in more difficult country.
In general it was found possible to work from one main camp for several days. Some general information was usually obtained about the probable direction of migration of the swarm, and the main camp was pitched some 30 miles away from the swarm along the probable route. From here the reconnaissance and morning observation parties would operate until the swarm had moved so far away from the camp that it would be difficult and wasteful in petrol and sleep to get to it for an early morning observation; when this became probable the main camp was moved to a more convenient position. Thus for the period 24th April to 8th May, only three main camps were necessary, one at Kericho, one at Eldoret, and one at Kitale.
The scientific party was thus occupied from an hour or two before dawn to about noon in making observations and in travelling to and from the roosts; some of the party would be further engaged in reconnaissance until after sunset, and no one was free from camp work in the camp. It was not therefore possible to make regular observations of swarm behaviour during migration; such observations as were made are dealt with under the headings of migration (p. 43) and settling (p. 52).
Fig. 1 Map of part of the Kenya Highlands, showing roosting places of a large swarm of locusts followed from Sotik (25/4/45) to the Turkwell Gorge (11/5/45). Each straight line joins roosting places and does not indicate the line of migration during the day. The roosting places are numbered to correspond with the text and the Tables. The swarm roosted at a nearly uniform altitude between Nos. 8 and 23. See pp. 8, 51.
go to next section of this publication