70 



Report S.A.A. Advancement of Science. 



absolute pressures over South Africa. The isobars on the following 

 map are founded on observation as far as the coast is concerned, but 

 on theorv across the continent. 



i"}}, 



Fig. 1. — Mean Aiiuual Isobars, S. Africa. 



The diurnal variation requires either hourly observations or 

 continuous records. I believe that hourly readings are taken at one 

 or two of the ports in Portuguese territory, but they are not published. 

 The only records available are those made by our indefatigable South 

 African meteorologist, Mr. J. R. Sutton, at Kenihvorth, near 

 Kimberley, and those made at the Johannesburg Government 

 Observatory. Some sixty odd years ago, hourly observations were 

 taken at the Cape and St. Helena, — these are referred to later on. 

 Hourly readings taken from the continuous records of Richard or 

 similar barographs, and checked against two daily readings of 

 mercurial barometers, are badly wanted for more places south of the 

 equator. 



The annual change of the barometer is the only element which 

 permits useful study at the present time, and even in regard to it, 

 there are difficulties. The barometers in common use in South Africa 

 are of the Kew pattern. This model suffers from several serious 

 defects ; one of these is the variable index error, which is apt to 

 change each time the barometer is moved, but a more important one 

 is that sometimes the mercury sticks and remains stationary for weeks 

 at a time. It is then necessary to invert the barometer and re-hang it 

 before it will perform properly ; a smaller defect is that the 



