l86 THE DESICCATION OF AFRICA. 



effective agents in producing rainfall over this area. The 

 direction and rate at which they travel will be dependent on the 

 currents some distance up : of these we know very little, except 

 that " cirrus " clouds travel from a little north of west. Now 

 the frequent occurrence of afternoon thunderstorms presupposes 

 an almost stagnant atmosphere, so tliat one is prepared for the 

 statement that the winds at Bulawayo are mostly light (forces 

 3 to 4 on the Beaufort Scale being considered strong for 

 Bulawayo) . 



The prevalent direction is very nearly E.S.E. Although 

 Father Goetz states that " it is not easy to decide whether there 

 are any w^inds especially associated with a fall of rain, he finds, 

 on correlating the self-recording anemometer and rain gauge 

 records, that 55 per cent of the total that fell between 24th 

 January, 1904, and ist May, 1908, occurred with winds between 

 E. and S., leaving 45 per cent, distributed among the remaining 

 five points and calms. 



If one may be allow^ed to apply these results to the Makari- 

 kari region, it seems reasonable to believe that the result of the 

 conversion of these marshes into a lake would be an increase 

 in the humidity of the air, particularly on the west side ; an 

 increased rainfall in the immediate vicinity of the lake ; and a 

 much smaller increase at a greater distance, noticeable all round 

 — possibly to a distance of two hundred miles — but more par- 

 ticularly on the north and west sides. 



In order to benefit the dry regions of the Union, a radical 

 readjustment of the distribution of pressure over South Africa 

 and the surrounding seas would be necessary, including the 

 obliteration oif the permanent anti-cyclone off the West Coast, 

 one of the most effective factors in causing the rainfall to be low 

 over the western half of this country. 



Such an alteration is certainly far beyond what could be 

 effected by Professor Schwarz's scheme,, large though one may 

 consider it. 



For these reasons I am afraid that tlie climatic results that 

 would ensue from the carrying out of this scheme would fall far 

 short of those depicted ; our arid regions would remain arid, and 

 the Karroo continue to be merely " Karroo."* 



* As mention has been made of Lake Michigan in the above remarks, 

 the subjoined extract from the letterpress of the U.S. Weather Bureau 

 publication, "' Meteorological Chart of the Great Lakes/' by Alfred J. 

 Henry and Norman B. Conger (1907), ought to prove of some interest. 

 These five Great Lakes together cover an area of about 93,000 square 

 miles, and lie directly in one of the main storm tracks of the North 

 American Continent, consequently the atmospheric circulation is mainly 

 horizontal. The prevailing winds are from north-west to south-west, so 

 that the effects of evaporation from the Lakes ought to be most apparent 

 to eastward. On the subject of "The Lake Influence," the authors 

 remark as follows : — 



" It is a rather common belief that the Great Lakes, by reason of 

 the enormous possibilities of evaporation they possess, are an effective 



