172 Transactions of the South African Philosophical Society. 



somewhat to — 3° - 5 at 7 p.m., after which it increases again. Thus, 

 on the whole, the effect of a depression is to exaggerate the diurnal 

 curve of temperature somewhat on the first day, rather less on the 

 second day, and to depress it on the third ; that is to say, in the 

 front of the depression the range of temperature is increased, in 

 the rear the range is decreased, while along the trough it is not 

 greatly affected. When we come to consider the variation in the 

 aspect of the sky we shall find, perhaps, a reasonable explanation 

 of the more rapid fall of temperature after sunset on the third day 

 in the decreasing cloudiness and vapour-tension, and consequently 

 increasing radiation. But the rapid fall after sunset on the first day 

 is certainly not so easily explained. It has undoubtedly some connec- 

 tion with the variation in the velocity of the wind, to be presently 

 referred to ; but the nascent clouds proper to this part of the 

 front of the depression, so far from acting as a blanket and 

 checking the escape of heat from the lower air, would seem the 

 rather to be actually themselves radiating cold to the earth. 



The behaviour of the dew-point is not in the least like that of the 

 temperature or pressure. Only from sunset of the first day to noon 

 of the second is the deviation in the quantity of water vapour from 

 the normal positive. The actual maximum positive deviation is at 

 sunrise on the second day. Throughout the preceding 30 hours 

 the quantity of vapour gradually increases relatively to the normal 

 diurnal curve, hour by hour ; in the following 42 hours the quantity 

 as regularly decreases, saving one or tw T o minor fluctuations not of 

 any great consequence. The semi-diurnal oscillation of vapour- 

 tension, of which the maxima come at about 10 a.m. and just before 

 sunset, is on the whole more disturbed than is the case with the 

 curves of the other elements. The times of the phases of the curve 

 of vapour-tension in the depression will be best understood by 

 reference to Table 6. The sunset maximum on the second day 

 seems almost entirely smothered. We learn also from the Table 

 that there is no great accession of water-vapour at any time during 

 the passage of a depression, the greatest positive deviation from the 

 normal tension not at any hour materially exceeding -015 inches. 

 In fact it is characteristic of a Kimberley cyclonic disturbance 

 that such variation as there is in the quantity of water vapour 

 obtrudes itself more in the dryness following the trough than in 

 the prior dampness. 



The deviation from the normal amount of cloud, though somewhat 

 irregular, as might be expected, is similar to that of the quantity of 

 water vapour. Starting from about the ordinary average at the 

 opening of the first day, the deviation is altogether positive during 



