RESEARCH IN BRITISH WEST AFRICA. 309 
of Lake Chad. The result during these months is an enormous fall of tempera- 
ture, and in the extreme north, when the wind, without having absorbed any 
previous moisture, meets the mists and vapours of the water systems 
of the Northern Territories, the temperature occasionally falls nearly to 
freezing point. 
Towards the end of the dry season, which varies in the different latitudes, the 
the tornados begin. These are cyclones from the north-east, accompanied by 
thunder-storms and torrential rains. They gradually merge into steady rains 
which last from July to October. During the rainy season, the atmosphere is 
laden with moisture, and a damp heat results, but for the rest of the year the 
Harmattan and a total absence of rain render the air excessively dry. Generally 
speaking, the nights are cool for the greater part of the year. 
Owing, however, to the enormous contrast in the various climatic conditions 
in different regions of the Protectorate, it is impossible to make any satisfactory 
generalisations. Take, for example, the case of Geidam and Ankpa in 1909. 
At the former station, the maximum and minimum temperatures recorded were 
110° Fahr. and 46° Fahr., while at the latter they were respectively 96° Fahr. 
and 56° Fahr. At Geidam, the total rainfall for the year was only 21°28 inches, 
while at Ankpa it was 66°85 inches ; at the former the rainy season was prac- 
tically confined to the months of May to September, but at the latter it extended 
from March to October. At Geidam, the maximum and minimum degrees of 
humidity were respectively 81 and 24 with a difference of 57, while at Ankpa 
they were 85 and 68 showing a difference of only 17. At the former, only 
during the month of August was there over 5 inches of rain (7°70 inches), while 
at the latter the following are the records—April 8°63 inches, May 14°02 inches, 
June 7°72 inches, August 8°12 inches, September 10°23 inches, October 
8°80 inches. 
With such an enormous difference in the climatic conditions in various parts 
of the Protectorate, it is not surprising to find that not only do certain species of 
blood-sucking flies show distinct local modifications, but there may even be a 
marked difference in the species which occur in the different districts. 
In order to see to what extent these are correlated, the following tables have 
been drawn up :— 
Table A :—The annual rainfall for various stations for the years 1906-1910. 
Table B:—The monthly rainfall for the same stations for the years 1909 
and 1910. 
Table C:—The maximum and minimum humidity at these stations from 
1906 to 1910. 
Table D:—The monthly temperatures for the same stations for the year 
1909. 
As these tables show graphically the main factors with which we are here 
concerned, I do not propose to deal with each at any length, but only to draw 
attention to the major features, 
