The Winds of Kimberley. ) 91 
moisture, and cloud, corresponding to the various deviations from 
the normal wind directions. The first three of these are for the 
hours 2 and 8 (a.m. and p.m.), but there are no observations of cloud 
at 2a.m. In the present section the instantaneous wind directions 
at the hours named are used. 
1. PressurE.—The method of reduction first attempted was the 
very simple one of adding together every pressure under assigned 
wind directions, and taking the mean for the whole three years. The 
resulting curves, however, were extremely irregular. The reason 
was not at first obvious, but a second examination of the work 
showed that certain directions happened to prevail say in a winter 
month with its customary high mean pressures, and other directions, 
say in a summer month with its customary low mean. To avoid 
this element of confusion, the process was adopted of drawing 
separate curves for each of the four hours of every month, and then 
reducing each curve by the simple addition or subtraction of a 
number, constant for that particular curve, to the mean pressure of 
the three years, the assumption being made, and supported to some 
extent by the monthly results, that each monthly curve resembled 
every other in shape. In this manner the effect of a greater or less 
number of directions from any assigned compass-point was elimi- 
nated. Finally the results were ‘bloxamed”’ in sets of three. 
Table 16 gives both smoothed and unsmoothed values. 
The mean pressures of the day have a primary maximum with 
winds from the south-east, and a secondary maximum with those 
from the north-west; a primary minimum with winds from the 
west-south-west, and a secondary minimum with those from. the 
north-north-west. There appears to be, as well, a tendency for 
the wind accompanying the highest pressure to radiate from a point 
which vibrates to and fro along a small are of the horizon in the 
course of the day. 
2. TEMPERATURE.—Table 17 gives the temperature results calcu- 
lated in the same way as the last. The mean temperatures exhibit 
one maximum with winds slightly to the north of north-north-west, 
and one minimum nearly south-south-east. But while the minimum 
temperature seems to accompany the wind of greatest air-pressure, 
the maximum travels with the sun (see the smoothed values), through 
pretty well a right angle. If future inquiry prove this to be a 
general law the passage of the warm point back to the north-north- 
east between 8 p.m. and 2 a.m. must be very rapid, and its hourly 
rate well worth study. 
3. Dew-pornts.—The dampest point is nearly midway between 
north-east and north-north-east, the driest point being exactly 
