MEASUREMENT OF RAINFALL. 85 



Wood, ^J.Sc, of the I'liion Observatory. In his paper on " The 

 Intensity Distribution of Rainfall over tlie Witwatersrand."' 

 abstracted in the Xatal Report (1907) of this Association, he 

 says : — 



A further comparison table shows that only _'o per cent, of the 

 Transvaal rainfall falls in what may be classified as light, directly beneficial 

 showers, against, roughly, 50 per cent, of the English rainfall; whilst about 

 60 per cent, of the Transvaal rain falls in heavy showers against only 25 

 per cent, of the English rainfall. It is, therefore, inferred that the greater 

 part of tlie Transvaal rainfall is not directly beneficial to the soil, nnd, 

 unless conserved, is lost to the Colony. 



The Burghersdorp weekly charts show that rain fell for 

 322 hours in 144 days during the same period, producing a 

 total precipitation of 63.55 inches, of which 9.22 inches fell in 

 207 minutes, z'i.:;., 14.5 of the total precipitation in 170 of the 

 total time. A sample of both charts is annexed hereto, on each 

 a two-hour continuous rainfall of similar intensity ha? been 

 marked. On the 24-hour chart it is a simple matter to deter- 

 mine the character of the fall for every five minutes, but on the 

 v;eekly chart it is very difficult" to do more than tell the total 

 precipitation for the period. Obviously, therefore, the 24-hour 

 chart is of a greater practical value than the weekly one. 

 Although both instruments above referred to have not been 

 calibrated with any standard instrument, the figures stated 

 herein mav be accepted as sufficiently accurate for the purpixse 

 of this paper. 



In advocating the more extensive use of the 24-hour auto- 

 rain gauge, it shculd be noted that the ordinary gauge has to 

 be read once every day, so that no extra work is entailed upon 

 observers b}- substituting an auto for the ordinary gauge, and as 

 already pointed out, the instrument is so simple in management 

 that no difficulty should be experienced in its continuous work- 

 ing, while, on the other hand, charts give better information at 

 headquarters than the ordinary cards. 



Finallv, a field of investigation lies in the direction of deter- 

 mination determining such problems as : — 



(a) What intensity of rainfall may be rated as non-bene- 



ficial to the soil for a given catchment. CIt is recog- 

 nised that this question is governed by many factors : 

 still, it is confidently expected that such an investiga- 

 tion would result in the establishment of some useful 

 standard.) 



(b) A\'hat intensity of rainfall may be regarded as phenom- 



enal, the disastrous results of which are often re- 

 ferred to as an " Act of God." 



(c) What intensity of rainfall should a designing engineer 



provide for in schemes to deal with flood water, and 

 at the same time keep within reasonable economic 

 bounds. 



These and many other questions are continually presenting them- 

 selves whenever the problem of the control of stormwater is 



