32 On the Determination of the Disgregation of a Body. 



gases; and he adds, "Although there exist bodies which do 

 not yield to the means at our disposal, we are nevertheless jus- 

 tified in concluding, from all known experiments, that all bodies 

 converge, in proportion as their temperature is raised, towards 

 this condition of perfect gas ; and this is enough for our argu- 

 ment." 



But this passage does not correspond with his calculations. 

 In order that equation (7), by means of which equation (5) is 

 reduced to equation (8), may hold as a general expression, we 

 require to have 



dT 



not only at very high temperatures, but at all temperatures under 

 consideration. We must therefore, if we admit M. de Saint- 

 Robert's formula?, suppose that every substance passes at all 

 temperatures into the condition of a perfect gas when the space 

 afforded to it for expansion is sufficiently great. 



But there are many bodies for which it seems to me that this 

 does not hold good ; it could not be said, for instance, of a piece 

 of iron, quartz, or any other similar substance, that it is suffi- 

 cient to increase the space into which it can freely expand in 

 order to cause it to pass at low temperatures into the condition 

 of a perfect gas. 



Even such substances as water, carbonic acid, and other com- 

 pound liquids or gases, present greater difficulties than might 

 perhaps be supposed at a first glance. We know, more parti- 

 cularly by the beautiful experiments of M. H. Saintc-Claire 

 Deville, that these bodies can undergo dissociation by the action 

 of heat. This dissociation no doubt involves internal work. 

 Unless, therefore, we suppose that complete dissociation occurs 

 at all temperatures, when the volumes are very great, we cannot 

 assume that the equation 



dT 



is true at all temperatures. 



We see from this that the expression for -^ Z derived from 



equation (5) is not generally identical with the expression F 

 given by equation (6), but that it is only in particular cases that 

 these two quantities can be regarded as equal, which is just what 

 I said at the outset. 



I will venture to add, in conclusion, a few words upon another 

 subject. 



There is an essential difference between my views and those 



