290 Professor Forbes on the Determination of Heights 
accurate enough for most purposes. The minute subdivisions 
of Deluc’s, De Saussure’s, and Wollaston’s instruments, are 
quite unavailing, as I have found by using the instrument of 
the latter with every precaution. 
My barometer having been broken in the course of my 
journeys, I was glad to have recourse to the boiling point as 
a means of estimating (only roughly as I expected) some 
remarkable elevations not before measured. In several cases 
I had the advantage of comparing my thermometric boiling 
point with a barometer, and lately I resolved to discuss these 
observations empirically, without reference to any theory or 
tables, or previous observations. 
I first projected the barometric pressures in terms of the 
corresponding thermometric observations. These were the 
following :— 
Barometer re- 
duced toEnglish 
inches, and to 
32° 
BorLine Point. 
August Tacul 200°-10 23'154 
Tacul 200°°6 23°358 
St Bernard 199°:08 22674 
Prarayon 201°°58 23°893 
M. | Col Collon 195°-15 20°77 
<u 29,11 A.M. | Gressony 204-20 25148 
September 5, mM. |, Martigny 210°-12 28°489 
I obtained a curve, which resembled a flattish logarithmic, 
the barometric numbers appearing to be in geometrical pro- 
gression, whilst the temperatures varied uniformly. This 
recalled to me an idea which I had entertained some years 
ago, that the boiling point would be found to vary simply 
with the height, to which I was led from knowing Deluc’s 
formula; but the idea had since escaped me, or been post- 
poned to other occupations. Now, however, I projected the 
simple elevations of the points of observation (derived from 
the barometric pressures from the common tables for com- 
puting heights uncorrected for the temperature), in terms 
of the boiling points, as in Plate VIII, and I was gratified 
to find, that a straight line passed almost quite through the 
whole of them, shewing that the temperature of the boiling 
