Diurnal Meteorological Conditions at Kimherley. 147 



ditions as to temperature are so exceptionally variable, to say- 

 nothing of the hour or so difference in the duration of sunshine in 

 April and July. Certainly at Kimberley we have an entirely 

 opposite state of things, in that with the low dew-point and clear 

 skies of June and July the total range of the barometer is -08 inch, 

 whereas in December and January, with their greater rainfall and 

 cloud, the range is upwards of -10 inch. And the place stands on 

 no isolated peak, but on the most typical of table-lands in the 

 world. 



It is at any rate worth while looking a little more closely into this 

 question of the effect of moisture upon the behaviour of the baro- 

 meter. Table 22 gives the mean annual hourly values of dew-point 

 and humidity for the four years, computed from the simultaneous 

 readings of dry and wet bulbs by means of the Greenwich Factors.''' 

 Table 23 gives the monthly values of the same elements compared 

 with the monthly values under clear skies. Here again, as before, 

 the yearly average for clear days is taken as the mean of the twelve 

 months. We see that in every case, when the sky is clear, the 

 dew-point and humidity-ratio are both less than the mean — as indeed 

 might have been anticipated. It has not been thought worth while 

 in this essentially preliminary statement, for the present at any rate, 

 to go to the great labour of determining the mean hourly dew-points 

 for clear days only. 



In Table 24 the pressures have been arranged according to the 

 magnitude of the dew-point. A progressive decrease of pressure is 

 shown to accompany an increase of moisture. Also there are some 

 indications of a secular increase of pressure for low dew-points, and 

 a corresponding decrease for high ones. The range numbers at the 

 foot of the columns have been estimated after the approximate 

 elimination of this secular variation. Their periods are : — 



A. From midnight to the first minimum ; 



B. ,, first minimum to first maximum ; 



C. ,, first maximum to second minimum; 



D. ,, second minimum to second maximum. 



The net result of the Table is rather disappointingly negative. 

 Excepting in the first column, including the very driest conditions of 

 the atmosphere, we have no indications at all of any such aqueous 

 influence upon the barometer as the comparison of clear and cloudy 

 skies might have been supposed to suggest. It might, therefore, at 

 first, appear that the clouds act not because of the greater moisture 



* See Glaisher's Hygrometriml Tables, 7th Ed., 1885. 



