378 
Transactions of the Royal Society of South Africa. 
3. Notes on the Eainfall Maps. 
Some preliminary monthly maps of the rainfall of Cape Colony and the 
Orange Free State were drawn by Gramble and published in the Report of the 
Meteorological Commission for 1.886. As the observing stations were few in 
those days, and the records of no great length, there was necessarily a good 
deal of imagination in the maps, and they only show^ in a rough way which 
places are w^et and which dry on tlie whole in given months. Buclian con- 
structed some monthly maps of the rainfall of Cape Colony, the Orange Free 
State, and parts of Natal and the Transvaal, based upon his discussion of 
the rainfall of the ten years 1885-94. These were published by the Meteoro- 
logical Commission in 1897 and w^ere a marked improvement upon G-amble's. 
They have been reproduced in Bartholomew's ' Physical Atlas,' 1899, and in 
various other publications. Of the 278 stations for which Buchan had the 
records, however, only about a half extended to the full ten years, the rest 
running to anything from three years to nine. Knox, in 1911, published a 
series of elegant monthly maps of the rainfall of Africa in his ' Climate of 
the Continent of Africa.' These were based upon odds and ends of scrap- 
book information culled from all sorts of responsible and irresponsible 
sources ; and consequently the beauty of the maps is no guarantee of their 
accuracy. Fraunberger and Herbertson have dealt wdth African rainfall at 
large. Father Groetz published a detailed map of the annual rainfall of 
Southern Rhodesia in the Meteorological Report of that Colony for 1908.* 
C. Stewart prepared a very fine map of the average annual rainfall of the 
Union of South Africa, and another showing the seasonal distribution of 
rainfall, for the use of the Select Committee of the Senate on Droughts, 
Rainfall and Soil Erosion. 
The rainfall maps now given are intended to represent graphically the 
broad features of all the results shown by the rainfall tables. In principle 
the construction of such maps is easy enough. All we had to do is to draw 
a series of lines, called " isoliyets," each one of which shall pass through all 
stations having a given rainfall. Thus the isoyhet of 1 in. (equal to about 
25 mm.) will join up in a continuous curve all stations in some particular 
area whose mean annual (or monthly) rainfall is 1 in. In actual practice 
it is not so easy to draw these lines as it is to state the way to do it. To 
begin with, very few stations out of the wdiole number can be found whose 
annual (or monthly) rainfall is some given quantity. It becomes a question 
not so much of joining up places having a given rainfall as of drawing 
an isohyet so that greater falls are on one side of it and lesser falls on the 
other. Next we have areas in which the rainfall varies very much over very 
small distances, as in the Cape Peninsula and in mountainous districts here 
* See also E. Goetz, "The Rainfall of Rhodesia," 'Proc. Rhodesia Sc. Assoc.,' vol. 
viii, 1909. 
