Some Statistics of TMwder and Lightning at Khnberley. 
107 
middle of each month. The first two digits stand for the hour, the second 
two for the minute. Thus, for January, 1854 " means 18 hr. 54 min. — 
that is, 6.54 p.m. 
The footlines in Table II are to be interpreted in the same way. The 
summer half shows one month (January, 1897) with no lightning observed. 
It is the first month of the observation period ; and consequently there may 
be some reason to question the vigilance of the lookout. March again 
leads the list with a maximum duration amounting to 62-2 hours (in 1899) 
and a minimum of 7"5 hours (in 1915). 
Kimberley thunderstorms are not confined characteristically to any state 
of barometric pressure, high or low. Statistics for five years give the 
following comparative numbers of these storms : 
Barometer Barometer Barometer 
high. mean. low. 
January to March ... 55 . 24 . 50 
April to June .... 5 . 3 . 20 
July to September ... 8 . 8 . 19 
October to December ... 43 . 23 . 47 
Year .... Ill . 58 . 136 
Hence in winter the few there are occur mostly with low pressures, 
whereas in the summer the barometric height is immaterial. What 
generally does happen in conjunction with the typical afternoon thunder- 
storm is a greater than normal fall of pressure from the morning maximum 
to the afternoon mimimum, as well as considerable and rapid variation 
while the storm is in progress. 
Since the bulk of the rain on the South African Table-land falls in 
thunderstorms, Table lY is added for purposes of comparison. It gives the 
number of rainfall hours in the same observation period, 1897 to 1919. 
This is not to be interpreted as duration in the strict sense ; but as the 
number of term hours in which rain was recorded, irrespective of whether it 
rained for each whole hour continuously, or only for a few minutes in each 
hour. Table lY agrees in its general features with Table I, albeit the 
turning points of its annual curve lag somewhat after those of the weighted 
annual curve of thunderstorms. There is a definite maximum of frequency 
about 4 p.m., and a definite minimum before noon. March has the highest 
frequency, August the lowest. The highest frequency in any one month goes 
to January, 1898, with 116 hours, for nearly 8^ in. of rain. During 
the whole observation period only one month (November, 1897) in the 
summer half had no rain at all, and only nineteen in the winter half 
had none. The statements one often sees in South African newspapers : 
" no rain for more than a year," " no rain for the last eighteen months," and 
so on, are untruths, although the cycle makers find them useful. 
