BY THE BOILING POINT OF WATER. 413 



a and C constants. But the first side of this equation is the very form which gives 

 elevations in terms of the barometric pressure. Hence the boiling temperature 

 varies as the height. In other words, the pressure varies in a geometrical ratio, 

 when the temperature of boiling, water varies uniformly ; but the pressure varies 

 geometrically when the heights above the sea vary uniformly : hence the heights 

 vary uniformly with the boiling temperatures. 



It is very singular that so elegant and simple a result should have escaped 

 every writer on the subject (so far as I know) ; even DELUC himself who proposed 

 the logarithmic law, and WOLLASTON, who unawares adopted the true law as a 

 first approximation, and then took a wrong one.* 



It is not to be supposed that the coincidence appears close, because the obser- 

 vations are not accurate enough to test it. Of seven observations between 195 

 and 210, no one differs 20 feet of elevation from the mean line, a quantity cor- 

 responding to 2 3 g of a degree, an amount which can by no means be considered as 

 being beyond the errors of observations ; and the small errors are well distri- 

 buted 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 upwards. In short, as is well known, the tensions 

 of steam cannot be represented by a geometrical progression in terms of the tem- 

 perature ; 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 con- 

 clusion. Indeed I cannot help thinking that the influence of the pressure 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 temperature can yield, and it is 

 another to ascertain under what pressure of air water will yield steam of a given 

 temperature. 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 indications as steady 



* He says, " Having occasion last summer of visiting Caernarvon, which would afford an oppor- 

 tunity of trying the instrument on the known height of Snowdon, and being aware that in 3550 feet the 

 variations of the boiling temperature were not to be considered uniform, as they might in small ele- 

 vations, on which alone I had before tried the experiment, I wished to provide myself previously with a 

 table for making the necessary correction, and from Dr UBE'S paper was supplied with data for calcula- 

 tion." Phil. Trans. 1820, p. 295. The table given from URE'S law of tensions gives a gradually in- 

 creasing number of feet, corresponding to every degree that the thermometer falls. 



VOL. XV. PAET III. 5 T 



