270 Transactions of the South African Philosophical Society. 
ever can be placed in the pressure returns so far published from that 
place, and the fortuitous agreement between the two methods only 
proved that the factor brought the level into very bad company. 
It is hard to say what the mean pressure at Bloemfontein ought to 
be, for at the beginning, even before the novelty should have worn 
off, the observing hours and seasons seem to have been erratic ; 
while throughout the “nineties ”’ the whole operation has approached 
the ridiculous. Latterly the results may have been not more than 
0-16 inch from the truth. 
It seems not unlikely that the factor 64250 will give fair altitudes 
above sea-level when the upper and lower stations are near the same 
latitude. Thus, for example, we know that the pressure increases 
very rapidly from East London along the coast to Durban: then 
taking 4,350 as the altitude of the Aliwal North barometer we deduce 
a 9a.m. mean sea-level pressure for the same latitude, say at Port 
Shepstone, of 30°125 inches. 
When it is not convenient to employ logarithms, Babinet’s 
excellent formula may be used with advantage.* It is— 
where Z, H, h, are as before, and C a numerical coefficient con- 
taining, like K, a number of petty factors. The approximate value 
of C, for the stations previously considered, is 55900, and the 
resulting altitudes are— 
Witaathartalys ee Uk Uae ts cede ener 2,354 feet 
Miliwall Niort ln ei eae te ee Senn rasta 4 300) ,, 
Philippolis seas eect eens ae 4,553 ,, 
Kimberley. 225, 2a assess eens eee 4,042 ,, 
Kenilworth ees ieee ke ee oe ea tee BIAS AOS | hays 
Passing now from generalities to details we see, according to 
Fig. 2, that high pressures are rare in the summer, nothing exceeding 
30:4 inches being at all likely at Durban from the end of November 
to the end of March, and that the high pressures, say exceeding 
30°66 inches, belong almost exclusively to the three months 
beginning about the end of the third week of July, extreme minimum 
pressures being also common at the same period. Sometimes we 
find low minima prevailing in conjunction with low maxima, or 
high minima with high maxima—suggesting here and there a certain 
sort of annual periodicity—and then again high minima accompany- 
ing low maxima, or vice versd, quite at random. Of the presumably 
* See Greenhill, Hydrostatics, p. 310. 
