A Gontrihutioyi to the Study of the Bain/all Ma'p of South Africa. 369 
up from quite a number of separate records fi'om here and there in a small 
area. One or two combinations of this sort have been done for East African 
stations, but some amount of caution will be requisite before accepting the 
results. 
Ill only one instance has any interpolation been attempted, namely, for 
the Devil's Peak lower gauge, in which an eleven years' record period, 1896 
to 1906, has been proportionally reduced in harmony with the Nursery and 
Blockhouse gauges to the thirteen years' period 1896 to 1908. 
With some slight modifications, the rainfall areas I to XVII are those 
used by the late Meteorological Commission of Cape Colony, and established, 
so to speak, under tliose numbers by many years of use. Some of these 
areas are large, and it has been thought best to subdivide them. Thus 
Section IX has been subdivided into IX A (West) and IX B (East). 
Sections XVIII to XXX are additional. 
The mean monthly falls for each section, or subdivision, have been 
computed, not necessarily by way of showing the average fall over the area 
(though it probably does show that with some degree of precision), 
but rather to trace the tendency of the monthly variations — which is the wet 
season and which the dry. 
2. Notes on the Rainfall Tables. 
I. Cape Peninsula. 
This small area, scarcely larger than the smallest English county, has a 
distribution of rainfall probably more varied than that of any other equal 
area in all Africa. From its position on the west side of the Continent it 
stands within a dry region, but the contour of the land alters the conditions, 
and transforms the dryness into an abundant precipitation over the greater 
part, which rises like an oasis out of an arid region. The stations in the 
table showing the least rainfall are Blaauwberg Strand (I, 2) and Cape Point 
(I, 3), with annual averages of 13-94 in. and 1270 in. respectively, these 
two stations being some thirty miles apart, one near the extreme north 
and the other near the extreme south of the peninsula. But according to 
Sparrman it never rains at all on the flats immediately to the south of 
Table Mountain* — a remarkable fact if true, seeing that Wynberg (I, 19), 
on the edge of the flats, has no less than 44J in, a year. Along the Atlantic 
side of the peninsula the rainfall is only moderate, averaging about 20 in. a 
year, Signal Hill (I, 13) with 18-52 in. and Sea Point (I, 12) with 21 in. 
The east, i. e. the landward side, is, however, wet, ranging between 
Eondebosch (I, 10) 42-56 in., Platteklip (I, 8) 45-34 in., Bishop's Court 
(I, 1) 57-90 in. The effect of altitude is well shown by the Devil's Peak 
gauges, the lower gauge (I, 4) at something over 1000 ft. 47-02 in., the 
* ' A Voyage to the Cape of Good Hope," Second Edition, 1786, p. 34. 
