268 Transactions of the South African Philosophical Society. 
of altitude apart the gravity correction may be ignored for the 
time. The air-temperature factor is— 
We se i as (4 
1 
= G0 
in which T is the mean temperature of Durban, and ¢ that of 
Kimberley, the mean temperature of the intervening air-strata being 
supposed to be 4(T + 7), which happens not to be quite the actual 
case. But whatever the actual mean temperature of the intervening 
alr may be, an examination of the numbers in Table 385 will readily 
prove that the numerical value of the air-temperature factor for all 
the given places must lie between the lmits 1:0733 and 1:0797, 
numbers equivalent at their full effect to an error of about 
13 feet of elevation in excess or defect. Assuming this temperature 
factor to be also constant, we have yet to consider a further 
correction for diurnal range. Now the daily pressure values for 
Durban refer to observations at [X. and XV. N.M.T. (= mean time 
of 30° E.), the corresponding number for Kimberley referring to 
observations at VIII. and XX. C.M.T. (= mean time of 18° 29’ E.). 
That is, the Durban barometer was read at 9h. 4m. a.m., and 
3h. 4m. p.m., local time, and the Kimberley barometer at 8h. 25m. a.m. 
and p.m. local time. As previously stated, the Kimberley values 
in Table 88 have been corrected, by the aid of the Kenilworth 
horary pressures, to agree approximately with readings taken at 
9h. 4m. a.m. and 3h. 4m. p.m., Kimberley local time. It seemed 
proper to make such a correction, but since the mean differences at 
Durban between the pressures at IX. and XV. (N.M.T.) are only 
about three-quarters of those for the same hours (C.C.M.T.) at 
Kenilworth, it is possible that the increased accuracy is not in the 
same ratio as the extra labour. Obviously the correct method would 
be to reduce all the observations to the mean of the day, if we only 
had the observations. 
Again in the comparison between Durban, Umtata, Aliwal North, 
and Philippolis, the times have been IX. N.M.T. at Durban, and 
VIII. C.C.M.T. at the other places, which reduces to, say,— 
9h. 4m. a.m. Durban local time, 
Sh. 25m. a.m. Umtata 5 
8h. 17m. a.m. Aliwal North local time, 
8h. 11m. a.m. Philippolis “ 
Fortunately on the mean of the year the South African barometer is 
not changing much between VIII. and IX., and thus the correction on 
account of difference of local time is very small—less in fact than 
