154 SECTION PHYSICS. 



wrong on that point, but it is the view which I have taken for many 

 years, and I believe it is the true view. 



Here is a tube which illustrates some of these conditions. It is a 

 sulphurous acid tube, and when you heat the liquid it first boils, and 

 then changes into the intermediate conditions to which I have already 

 referred. Here is another tube, which I am almost afraid to open, not 

 because it contains carbonic acid, but because it is not so strong. It 

 was made by Professor Dewar, and contains liquid carbonic acid. 



I will very briefly refer to the recent investigations I have made on 

 the properties of matter in the gaseous state. There are three impor- 

 tant laws which are known with regard to the gaseous state : the law 

 of Boyle, according to which gases vary inversely in volume accord- 

 ing to pressure ; the law of Gay-Lussac, according to which they vary 

 in volume by a definite amount of their volume at zero, or any other 

 fixed temperature ; and the law of Dalton, according to which gases 

 in mixture act as if the other gas were not present, as if each gas 

 occupied the whole space. At high pressures these laws fail. They 

 are true in the ordinary conditions of gaseous matter approximately, 

 very closely indeed in the case of some of the gases, but when you 

 operate under these great pressures all these laws fail. The failure of 

 the law of Boyle has been known to some extent, but that of the law 

 of Gay-Lussac, I believe, has not been observed till my last commu- 

 nication to the Royal Society. 



I will conclude by writing down three formulas which express 

 ascertained facts. The expansion by heat of a body in the ordinary 

 gaseous state, whether measured by its expansion at constant 

 pressure, or by the increase of elastic force under constant volume, is 

 not a simple function of the initial volume or initial elastic force but a 

 complex function changing with the temperature. There are certain 

 points to which I have given the name of homologues. The following 



are the formulae p, = *> If you take any two isothermals, the value 

 of \i is the same throughout the whole range of isothermals. 



(l p V] V = C. 



