Pressure and Temperature Results for the Great Plateau. 269 
the possible secular error of a single observation with a marine 
barometer—although in the case of Philippolis it may amount to 
fully 0-008 inch during the winter months. Making the necessary 
corrections from the Kenilworth observations, allowing for the 
differences of local time, we get finally the values, for 9h. 4m. local 
time, of Table 39. 
Granting the cistern of the Kimberley barometer to have been 
4,042 feet above the sea, and substituting this value of Z in the 
formula, we derive K = 64250 approximately. If this result hold 
for South Africa generally, we determine respective solutions for 
LZ to be— 
WSininthantenhgnasecc ces eee saa reson SO ORE cele 3 2,352 feet 
Atlin ale INO iblay, 2s assed ee eee conn cutee een aL BIG | 5. 
alatiiqopOlisry scans aston con aabes alenicselosaees ome 4,906 ,, 
Here Aliwal North comes out 12 feet higher than the railway level, 
and 32 feet higher than the altitude published by the Meteorological 
Commission ; but Umtata is 48 feet less, and Philippolis 144 feet less 
than the Commission has said. It has never been explained how 
the altitudes given in the Annual Reports of the Meteorological 
Commission have been determined, but, since they are generally in 
round numbers, it is a fair inference that they claim no great 
excellence. Kenilworth (Kimberley) with the same factor, for some 
four years’ mean, comes out at 3,928 feet. 
Professor Guthrie, about twelve years ago, constructed a table of 
altitudes above Cape Town sea-level,* using K = 64300, and made 
a remarkable, not to say miraculous, agreement for quite a number 
of places all over the country between the altitudes by railway 
levelling and those by the barometer; the differences, for example, 
between the two methods being only 2 and 4 feet respectively at 
Aliwal North and Kimberley. He depended upon a mean pressure 
of 25°762 inches at Aliwal North, and 26-031 at Kimberley, at about 
8 a.m. local time. The former agrees closely enough with the value 
in Table 39 (= 25:777 inches at Yh. 4m. a.m.) ; but the Kimberley 
value is almost certainly 0:04 inch too low at least. Moreover, he 
has a ditference of only six feet between the barometer and the level 
at Bloemfontein, of all places! The observing is pretty feeble at 
many second-order stations in the world, and in South Africa 
especially, but if a worse lot of barometer results than that taken at 
the capital of the quondam “ Free State’”’ has ever been evolved, it 
ought to be carefully preserved—as a record. No confidence what- 
* F. Guthrie, ‘‘Sea Levels in South Africa from barometric observations ”’ 
(Trans. S. Af. Phil. Soc., vol. v., p. 318). 
