266 Transactions of the South African Philosophical Society. 
7. In August the same state of things prevails. For although the 
temperature rises slightly over the coast, the rise on the table-land 
is at least four times as great; and thus while the seasonal fall of 
pressure has well set in, over the whole southern hemisphere, it is 
checked a little on the coast and increased on the table-land by the 
outward expansion arising from the more rapid increase of tempera- 
tures. In September the pressure surfaces are rising generally, and 
the differences begin again to decrease. 
8. In October we see another increase in the difference of 
pressure. It is very plain at Aliwal North and Kimberley, and 
appears as a considerable temporary check in the curve rates of 
Philippolis and Umtata. Now at the coast stations the monthly 
rate of increase of temperature augments far into the summer, 
reaching its maximum from November to December. On the table- 
land the maximum rate is reached from August to September. The 
middle stations seem to feel the influence of both, having a marked 
break in their rates of temperature increase from September to 
October. Corresponding to this is the marked October decrease in 
the rate of fall of the barometer at Durban, where for pretty well a 
whole month the pressure curve les above the inverted Kimberley 
temperature curve. <A glance at the daily values in Tables 10 and 
22 will be sufficient to prove this month of high relative pressure an 
essentially Durban phenomenon: the rate of fall of pressure from 
September to October being considerably checked there, slightly so 
at Umtata, and not at all at the higher stations. The interpretation 
is not obvious. Perhaps it is not unlikely that the check to the rate 
of rise of temperature at the middle stations may also indicate a 
sufficient check to the rise of the pressure surfaces to allow the ex- 
panding air of the table-land—which experiences no such check—to 
again, temporarily, push outwards over the coast. The check of the 
temperature increase concerns the maximum almost entirely. 
The general trend of the changes just sketched in detail will be 
best understood by superimposing Fig. 2 upon Fig. 1; from which 
it will appear that the Durban pressure curve falls below the inverted 
Kimberley temperature curve in autumn, and above it in spring, 
whereas the Kimberley pressure curve is above its inverted tempera- 
ture curve in summer, and below it in winter.* Of course it will be 
understood that the above sketch aims at neither completeness nor 
great exactitude. That unfortunately is out of the question, and 
* Compare V.Bjerknes, ‘‘ The Dynamic Principle of the Circulatory Movements 
in the Atmosphere” (Monthly Weather Review, Oct., 1900). This paper came after 
the above sketch in the text was written: too late for use here. Iti of exceptional 
interest. 
