(61) 
EAINFALL AND THE PEESSUEE GEADIENT. 
By J. E. Sutton. 
The seasonal rainfall over the greater part of South Africa may be 
regarded as due mainly to a planetary rain belt following the sun to and fro 
across the equator. It is this that gives us the summer rains. As the belt 
moves north it is followed at a distance by the winter rains of the south-west, 
which, however, do not penetrate often so far inland as the middle reaches 
of the Orange Eiver, and which recede as the belt returns south. Over the 
central table-land of South Africa the seasonal rains are very unequal ; and 
the climate is to be regarded in general as a permanently semi-arid one 
punctuated at occasional intervals by wet years. Why the rainfall varies 
so much from year to year may be suspected from a consideration of the 
questions raised by Colonel Eawson in his notable papers on the movements 
of anticyclones. It appears to depend upon the variations of barometric 
pressure, or perhaps, rather, upon the causes which produce these. Father 
G-oetz has made an important advance in this aspect of the matter by his 
discovery of a relationship between the fluctuations of rainfall at Bulawayo 
and the variations of the second harmonic term in the annual pressure 
curves. 
The present paper is a brief discussion of the possibility that the pressure 
gradient between Kimberley and Cape Town may have some sort of connec- 
tion with the rainfall of the two places. For our purpose the monthly means 
of pressure and rainfall of the two places have been compared for the twelve 
years 1897 to 1908. The pressure gradient is supposed to be represented 
month by month by the simple arithmetic difference between the monthly 
means of pressure at the two places.* 
Table I gives the mean gradient, and the mean rainfall of each place for 
each month of the period. On the whole the gradient is least in autumn 
(March to May), and greatest in spring (October and November). There is 
a maximum of rainfall at Cape Town near midwinter, and a secondary maxi- 
mum near midsummer. The maximum rate of fall at Kimberley occurs in 
February, the minimum in the late winter, with some tendency to a 
secondary maximum in July. 
Table II shows the mean monthly rainfall of the two places for gradients 
less than the median (" Small Gradients 1 '), and for those greater (" Large 
Gradients "). 
(1) Cape Town. — At Cape Town, speaking at large, the rainfall when the 
gradient is small exceeds that when the gradient is large by about 18 per 
cent., both summer and winter. Eight months show the excess, four do not. 
* I.e. at the Royal Observatory, Cape Town, and the Observatory at Kenil worth 
(Kimberley). 
