The Winds of Kimberley. | 19 
this mere guesswork. Calms, which are comparatively rare, are 
included equally with the others. Some difficulty was experienced 
in assigning’ the direction in those cases where the mean hourly 
direction fell very nearly between two of the sixteen directions, and 
there seems to be no doubt that a certain amount of bias displayed 
itself from time to time favouring some of the points in question at 
the expense of those adjacent. It might be expected that such a 
bias, not being deliberate, would rectify itself in time, but the gradual 
accumulation of the numbers to the final totals did not completely 
bear out the expectation. At the same time it may be claimed that 
no error large enough to materially alter the true monthly or hourly 
‘resultants was introduced in this way. 
Table 1 gives the number of hours of wind, irrespective of velocity, 
for each of the sixteen principal compass points during each month 
from the three years’ observations. Fig. 1 is a graphical representa- 
tion of what may be called the resultant wind-direction calculated for 
each month, and for*the whole year, from the numbers of Table 1. 
It is formed by drawing rectangular axes NOS, HOW, through the 
origin O, and projecting the number of hours of each wind upon 
them. Thus each direction multiplied by its numerical coefficient 
will have two components, one upon the axis NOS, and the other 
upon HOW. All components falling along HOW in the direction 
OE are conventionally plus, and all in the direction OW minus. 
Also ON and O§S are in the same way conventionally plus or minus 
respectively. The final resultant is formed from the components by 
the principles of the parallelogram of velocities. Consider, for 
example, the wind numbers for January, and let N, E, S, W, repre- 
sent the respective components measured along ON, OH, OS, OW; 
R being the final resultant. Then we have— 
N = 185 + (227 + 144) cos 223° + (149 + 161) sin 45° + (107 -+ 195) sin 223° 
E = 97+ (195 + 130) co ope 4 (161 + 93) sin 45° + (144 + 110) sin 223° 
s= i + (173 + 110) cos 223° + (141 + 93) sin 45° + (138 + 180) sin 2230 
W = 59 + (107 + 138) co cos 223° + (149 + 141) sim 45° -— (227 + 173) sin 223° 
_N—S =228°1 
E—W = 30°6 
Whence es Jy (Ozer e cae 6)? | —930°14 
And tan es DOK 
The components for each month found in this way are given in 
Table 2. But it must be noted that neither the components, nor 
their resultants, of themselves are to be interpreted as necessarily 
