292 Professor Forbes on the Determination of Heights 
quantity corresponding to ,°, of a degree, an amount which 
cannot be considered as being beyond the possible errors of 
observations; and the small errors - are well distributed 
throughout. On the contrary, when the tensions of vapour, 
from Dalton’s Table, are projected beside them, as in the 
dotted curve of the figure, not only do they lie wholly above 
the line, but these tensions cannot be represented (when 
treated as representing barometric heights) as a straight 
line at all. They have a manifest curvature convex up- 
wards. In short, as is well known, the tensions of steam 
cannot be represented by a geometrical progression in terms 
of the temperature; but when water boils in the free air, 
the pressures are then exactly in geometrical progression. 
I never saw any ground for believing that the two laws 
must be the same. Our theory of vapours is not sufficiently 
perfect to admit of our drawing any such conclusion. In- 
deed I cannot help thinking that the influence of the pres- 
sure of the air upon the elasticity of nascent steam, is a fact 
not easily reconciled with Dalton’s theory of the pressure 
of elastic fluids. It is one thing to ascertain the elasticity 
of steam of maximum density, which water of a given tem- 
perature can yield, and it is another to ascertain under what 
pressure of air water wil] yield steam of a given tempera- 
ture. In practice I have observed the temperature of the 
boiling water, and not of the steam. The construction of 
the apparatus required this. But by moving the furnace to 
a side, so as to prevent the flame from disengaging the steam 
immediately under the thermometer, I have found the indica- 
tions as steady as I believe can be got in any other way. The 
mass of the water and also of the thermometer favours this. 
But I had a farther test of the exactness of the arithme- 
tical progression above established, and that as severe as 
could be proposed. It was to compare De Saussure’s obser- 
vations on Mont Blane, and the pressure there observed, with 
the result of my formula. But first, it was necessary to 
correct the zero point of his instrument, and to render it 
comparable to mine. De Saussure’s boiling point, 80° of 
Reaumur, or 212° of Fahrenheit, was adjusted at 27 French 
inches, or 28.777 English. 
