An Introduction to the Study of South African Rainfall. 3 



tank. These are generally too small as compared with the gauge, 

 for reasons to a large extent obscure. 



In consequence of the great local variation in the falls of rain, a 

 number of 5-inch Snowdon gauges were set up in 1897 at the different 

 mines and on some of the farms owned by the De Beers Company. 

 'The records were in many cases broken during the late war, and all 

 the outside gauges taken away by the Boers. These gauges were 

 :all placed at the same height above the ground as the Kenilworth 

 gauge, but they are mostly in a more open situation. This may be 

 partly the reason why their catch is so frequently less than that of 

 Kenilworth.* 



Tables 1-12 give a complete history of rainfall at Kimberley from 

 1877-1902. The daily amounts for 1877, 1888, 1889, and part of 

 1884, were kindly given me by Mr. Matthews; 1880-1883, the 

 remainder of 1884, with 1885, are from the Exploration Company's 

 register ; 1886-1897 are from the Lee register ; 1898-1902 are from 

 the Kenilworth register. Thus the history deals almost exclusively 

 with the records of 8-inch gauges,! one half giving the rainfall near 

 the ground, the other half at a height of 15 feet. 



It has been usual at Kimberley to reckon the fall up to 8 a.m. on 

 the last day of each month, entering the separate items to the date 

 •of measurement ; but because it is preferable to give the fall up to 

 •8 a.m. on the 1st of the following month, I have rearranged the 

 monthly totals accordingly ; and, so as not to disturb previous dates, 

 made a rain month of 31 days extend from the 2nd to the 32nd. 

 'The totals in the last column of the historical Tables give the 

 number of days, and the total fall for each date, in twenty-six years. 

 If annual averages are required they may be obtained after division 

 by 26. 



It appears from these Tables that March is the wettest and July 

 the driest month, the increase or decrease from one to the other 

 being gradual. Although the total fall is less in February than" in 

 January, February is really the wetter month by an average of 

 •07 inch per day, on account of the fewer days it contains. There 



* The student may with profit consult Heberden, " On the different quantities 

 ■of Eain which appear to fall at different heights, &c," Phil. Traits., vol. lix. 



p. 359 (1769); Barrington, "Experiments made ... to ascertain the different 

 •quantities of Kain ... at different heights," Phil. Trans , vol. lxi. p. 294 (1771) ; 



and chiefly, on the general question, Hann, Lehrbuch tier Meteorologie, p. 296 

 •Cleveland Abbe, " Determination of the true amount of Precipitation, &c," Bulletin 



No. 7, F. 1)., U.S.A. Dept. of Agr. ; also G. Dines, " Difference of Rainfall with 

 •elevation," British Rainfall (1880). The other literature on the subject is of 

 •enormous bulk, and often amazingly futile. 



f Not that the size of the gauge makes any appreciable difference. 



