466 
Proceedings of the Royal Society , 
ness of accuracy, and not the observations themselves. Since the 
tables which have been prepared for travellers are calculated on the 
assumption that 29*92 inches, or 300 inches, is the zero point for 
heights, there can be little doubt that, by this method, the heights 
of many plateaux and mountains of the globe have thus been 
determined. 
From my paper, read before this Society in March 1868, on the 
Mean Pressure of the Atmosphere over the Globe, illustrated with 
three charts, showing the Mean Isobaric Curves for July, January, and 
the year, it may be seen that a pressure of from 29*9 to 30'0 inches 
is very near the mean annual pressure over the greater part of the 
globe, particularly over those portions of it explored by travellers. 
But when we examine the months, it is at once apparent that 29'9 
inches is very far from the mean pressure in many regions. This 
point will be illustrated by the pressures at Barnaul, Siberia, which on 
an average of 19 years are, reduced to 32° and sea-level, as follows : — 
Mean atmospheric pressure at Barnaul in July, 29*536 inches. 
„ „ „ January, 30-293 ,, 
„ „ „ Year, 29'954 „ 
Suppose, now, it be proposed to ascertain the height of Lake Balkash 
on some day in July, the pressure at the time being the average 
of the month. Let the observed pressure be 28 - 8 inches reduced to 
32° F., and the temperature of the air be 70 o, 0, then if the sea-level 
pressure be assumed to be 29-9 inches, it is plain that the difference 
due to height is 1*10 inches; in other words, the height of the lake 
would be, in round numbers, 1080 feet. But since the sea-level 
pressure of this locality, which is nearly that of Barnaul, is 29*536 
inches, the difference of pressure due to height is only 0-736 inch; 
the height, therefore, is only about 730 feet. Again, if in January, 
when the barometer is the mean of the month, the pressure at Lake 
Balkash was observed to be 29-42 inches, and the temperature of the 
air 1°*0, assuming that 29*9 inches is the mean sea-level pressure of 
January, 0*48 inch is the difference of pressure due to height — thatis, 
the lake is about 400 feet above the sea. But since the mean pressure 
is nearly 30-3 inches, 0‘88 inch is the pressure due to height; the lake 
is therefore nearly 730 feet above the sea. Thus in July the lake 
would be made 350 feet too high, and in January 330 feet too low — the 
