469 
of Edinburgh, Session 1868 - 69 . 
to a pressure of 27231 inches.* But since the observation was 
made about 9 a.m., when the pressure is about the maximum of the 
day, subtracting *043 inch as the correction for daily range in 
July, we obtain as the mean pressure of the day 27*188 inches. 
If we assume the sea-level pressure to he 29-70 inches, the differ- 
ence due to difference of height will be 2-512 inches, and the tem- 
perature of the air being at the time 84-0, the height of Albert 
Nyanza will be in round numbers about 2550 feet, or considerably 
under the height usually given. 
Similarly, by the same reasoning, Gondokoro, calculated from Sir 
Samuel W. Baker’s observations to be 1999 feet in height, will be 
only about 1800 feet above the level of the sea. 
Considering the small difference within the tropics in the mean 
pressure of any month, say July, from year to year, it follows that 
if recent African travellers had been provided with good ther- 
mometers for determining the boiling-point of water, and had 
made carefully conducted observations with them, noting the 
precise hour and month of the observations, one of the great 
problems of African travel would have been already solved, viz., 
whether Lake Tanganyika does or does not flow into Albert 
Nyanza, unless the difference of level between these two lakes is 
comparatively small. But since travellers have been given to 
understand that the heights deduced from their observations may 
be in error to the extent of from 300 to 500 feet, less care has 
been bestowed in making such observations than would otherwise 
have been the case. 
In extra-tropical regions the height of the barometer is much 
more fluctuating, and the pressure during any month from year 
to year varies more than within the tropics. But even in these 
regions the limits of error are much less than are usually supposed, 
if care be taken to make the observations full and precise, so that 
when they come to be reduced it may be in the power of the 
meteorologist to value them at their proper worth. This remark 
may require a little explanation. 
In temperate regions barometric fluctuations are more frequent 
and of greater amplitude in such countries as Great Britain, which 
are situated between a continent on the one hand and an ocean on 
* Regnault’s Tables, revised by Moritz. 
VOL. VI. 3 P 
