Pressure and Temperature Results for the Great Plateau. 267 
will so continue until meteorology is better served by the Legislature. 
There are chances here, too, for kites and balloons. Nor is it quite 
clear why the Kimberley wind-vane tends to make one complete, if 
at times vacillating, clockwise rotation per annum. That one 
depends to some extent and in some way upon the other is a reason- 
able idea. Certainly the Kimberley diurnal wind stratum is very 
shallow, and possibly the annual stratum is so also. The seasonal 
transfer of air from one hemisphere to the other may take place at a 
great altitude, and the transfer from coast to table-land and back 
somewhere between, in the middle reaches. The barometer neces- 
sarily integrates the combined effect of the three movements. 
The chances of solving the problem of reducing the table-land 
pressures to sea-level are not simplified by our results so far, 
although the conditions may be made clear by future research along 
similar lines. And the difficulty exists not merely because, as is 
commonly stated, our knowledge of the temperature of the air is 
limited and fragmentary, but because the ordinary formula is insuffi- 
cient. But at the least the annexed tabular matter ought to say to 
what extent it is likely to be so. Jt is a question whether it would 
not be possible, if we had the material, to make a synoptic weather- 
map for the higher stations reduced to, say, a 4,000-foot datum, and 
another for the coast and middle districts reduced to sea-level, com- 
bining the two by some suitable process afterwards. On the whole, 
however, it is likely that a weather-map depicting only proportional 
departures from the normal daily means will be the most useful tool 
of our future weather-bureau—when we get it ! 
Although the average monthly differences of pressure between 
Durban and Kimberley, deduced from two observations per diem, 
may not be so exact as they might be if they were deduced from 
one observation per diem for twice the number of years, yet for the 
annual mean difference they are doubtless quite as good. Assuming 
this much, we may calculate the values of the numerical coefficient 
in the formula— 
7 Jal 
Li OS, 7 
in which Z is the altitude of Kimberley in feet, H is the mean 
height of the barometer at Durban sea-level (mean low water), and 
h the mean height of the barometer at Kimberley, H and h being 
corrected and reduced to 32° F. K includes a number of petty 
factors; the gravity correction for both altitude and latitude, correc- 
tions for air-temperature, and for humidity. Since Durban and 
Kimberley are only about a degree of latitude and some 4,000 feet 
—————————— —— 
