4 Transactions of the South African Philosophical Society. 



seems, however, to be a distinct tendency to wet and dry seasons 

 characteristic of certain seasons. For example, the first half of 

 February and the second half of November are notably deficient in 

 quantity ; also the latter half of June, the latter half of September,, 

 the fourth week of October, and the first half of December, are rela- 

 tively wet. The absolutely wettest time is during the last week of 

 February ; the driest during the first week of August. 



Of the accidental variations we have some remarkable cases, due. 

 chiefly (probably) to the shortness of the record. A good example is. 

 found in the twenty-four hours ending at 8 a.m. February 9th. In. 

 the whole twenty-six years there was not a single shower on this, 

 date so great as *10 inch, and indeed only one shower of any import- 

 ance on the 10th ; the total fall for the two dates together in the- 

 whole twenty-six years only reaching three-quarters of an inch. But 

 as it happens the rule for dry weather at this time breaks down in 

 the present year, 1903, with "45 inch — a greater fall for the date 

 than is furnished by all the former twenty-six years put together- 

 Still it is curious, to say the least, that February 9th is the driest 

 day, and February 9th and 10th the driest two days, in sixteen con- 

 secutive weeks. On the other hand, the rule for each of the four- 

 dates, February 25th-28th, is for heavy rain in every other year. 

 Dates upon which no rain at all is known to have fallen are only 

 found between June and September ; albeit too much stress should 

 not be laid upon the apparent dryness or otherwise of particular- 

 dates in any record extending over less than fifty years, because, for 

 one reason, the total rain as given for any date is not materially 

 greater than the possible rain in twenty-four hours. Take, for 

 example, the rain of September 17th. This, for at least a quarter of 

 a century, from 1877-1901, seemed to be one of the driest of days, 

 only one small fall of *015 inch being on record. In 1902 came a 

 change with the heaviest day's rain known here [i.e., 4-52 inches),, 

 making September 17th at one jump the wettest date in six months.. 

 This is a heavier fall than some rainfall maxims allow for. For 

 example : " The extreme percentage of the mean annual rainfalL 

 which may be expected on any one day is, with a mean fall of 

 20 inches, 16 per cent. ..." * For the same mean annual fall 

 E. M. Eaton gives 20 per cent, f The Kenil worth fall in question is, 

 however, nearly 25 per cent, of the annual mean. Moreover, in 

 India there are stations where the fall in one day has exceeded the; 

 annual mean.]; 



* British Rainfall, 1867. 



f Symdns's Met. Mag., 1878, p. 13. 



I See Blandford, " Climates and Weather of India," pp. 64, 266.. 



