Variation of Hourly Meteorological Normals at Kimberley. 175 



direction of a cyclonic wind in the rear of a depression. Thus we 

 get accelerated velocities after sunset of the second day, and, by the 

 same token, retarded velocities after sunset of the first day. But 

 such a line of suggestion leave unexplained, and in fact contradicts, 

 the low velocities before sunrise of the second day and the high 

 velocities before sunrise of the third. The rapid slackening of 

 velocity after sunset of the first day relatively to the normal curve 

 deserves attention if only because of the rapid cooling of the air at 

 the same time. 



Table 11 gives the total number of hours of wind from each of 

 16 directions in the 105 depressions (i.e., 315 days) we are consider- 

 ing. There seem to be two prevailing directions on both the first 

 and second days. On the first day directions in the first and second 

 quadrants prevail ; on the second day in the second and third 

 quadrants. On the third day there is a single prevailing direction 

 in the third quadrant ; so much so that the wind is more peristent 

 from the south-south-west than from all points having a northerly 

 component, including east and west, put together. The actual 

 diurnal variation of each direction is not materially affected at any 

 time, e.g., the third quadrant winds have their maximum frequency 

 near the same hour of the afternoon on all three days as under 

 ordinary circumstances. 



The hourly components of wind-frequency are set out in Table 12. 

 The north component commences the three days slightly negative 

 (i.e., it is slightly southerly), but changes to positive about 1 a.m., 

 .and remains so until the trough has passed, after which it is negative 

 for the remainder of the period. It has positive maxima at 10 a.m. 

 on the first day, and at 9 a.m. on the second, while the negative 

 maxima fall at 7 p.m. on the first day, 6.30 a.m. and 7 p.m. on the 

 third. The east component has positive maxima about an hour 

 before sunrise, and negative maxima about 2.30 p.m. on each day. 

 Thus the phases of the east component are practically unaltered 

 by the depression, such changes as the Table reveals falling 

 almost entirely upon the north component. This very important 

 fact should be compared with a certain previous result that the east 

 component curve at Kimberley is essentially a curve of temperature. 



The angle <f> in Table 12 represents the angle described by the 

 resultant wind direction from the east round by north, west, 

 south. This angle is greater than the normal -between midnight 

 and 10 a.m. on the first day, and between midnight and 3 p.m. on 

 the second day, but it reaches its starting position on the second day 

 not by boxing the compass but by backing after 4 p.m. on the first 

 day. Throughout the whole of the second day the resultant veers 



12 



