24 REPORT— 1902. 



certainly gave remarkable support to Amontons' notion of the zero to 

 find that simple gases like hydrogen and compound gases like ammonia, 

 hydrochloric, carbonic, and sulphurous acids should all point to substan- 

 tially the same value for this temperature. But the most curious fact 

 about this research of Clement and Desormes is that Gay-Lussac was a 

 bitter opponent of the validity of the inferences they drew either from 

 his work or their own. The mode in which Gay-Lussac regarded the 

 subject may be succinctly put as follows : A quick compression of air 

 to one-fifth volume raises its temperature to 300 degrees, and if this could 

 be made much greater and instantaneous the temperature might rise to. 

 1,000 or 2,000 degrees. Conversely, if air under five atmospheres were 

 suddenly dilated, it would absorb as much heat as it had evolved during 

 compression, and its temperature would be lowered by 300 degrees. 

 Therefore, if air were taken and compressed to fifty atmospheres or 

 . more, the cold produced by its sudden expansion would have no limit. 

 In order to meet this position Clement and Desormes adopted the 

 following reasoning : They pointed out that it had not been proved that 

 Gay-Lussac was correct in his hypothesis, but that in any case it tacitly 

 involves the assumption that a limited quantity of matter possesses an 

 unlimited supply of heat. If this were the case, then heat would be 

 unlike any other measurable thing or quality. It is, therefore, more 

 consistent with the course of nature to suppose that the amount of heat 

 in a body is like the quantity of elastic fluid filling a vessel, which, while 

 definite in original amount, one may make less and less by getting nearer 

 to a complete exhaustion. Further, to realise the absolute zero in the 

 one case is just as impossible as to realise the absolute vacuum in the 

 other ; and as we do not doubt a zero of pressure, although it is unattain- 

 able, for the same reason we ought to accept the reality of the absolute 

 zero. We know now that Gay-Lussac was wrong in supposing the 

 increment of temperature arising from a given gaseous compression would 

 produce a corresponding decrement from an identical expansion. After 

 this time the zero of temperature was generally recognised as a fixed 

 ideal point, but in order to show that it was hypothetical a distinction 

 was drawn between the use of the expressions, zero of absolute tempera- 

 ture and the absolute zero. 



The whole question took an entirely new form when Lord Kelvin, in 

 1848, after the mechanical equivalent of heat had been determined by 

 Joule, drew attention to the great principles underlying Carnot's work 

 on the ' Motive Power of Heat,' and applied them to an absolute method 

 of temperature measurement, which is completely independent of the 

 properties of any particular substance. The principle was that for a 

 difierence of one degree on this scale, between the temperatures of the 

 source and refrigerator, a perfect engine should give the same amount of 

 work in every part of the scale. Taking the same fixed points as for the 

 Centigrade scale, and making 100 of the new degrees cover that range, it 

 was found that the degrees not oidy within that range, but as far beyond 



