ON THE FLOW OF SOLID BODIES. 257 



thermo-dynamical fact. If this fact had never been observed before, it 

 was clearly owing to the circumstances necessary for its manifestation 

 never having been all united together in so favourable a manner. 

 Brightened platinum requires for its deformation a great amount of 

 work. Its surface is unalterable and almost translucent when the metal 

 is heated to a red hot temperature ; it is but a moderately good conduc- 

 tor of heat ; its calorific capability is somewhat feeble ; all conditions 

 rendering the phenomena visible in the smelting of platinum whilst it 

 has passed unnoticed in the case of all the other metals. 



But, although anticipating this explanation, we, nevertheless, felt it 

 our duty to prove it by more direct experiments, of which I shall now 

 speak, and which constitute the chief novelty may I say the chief 

 interest of this communication. 



Given a metallic bar at the ordinary temperature if two of its 

 lateral surfaces be coated with wax or tallow, and is tnen subjected to 

 the action of a single stroke of a ram, the wax will melt opposite the 

 depression which is produced ; and it is proved that, in certain cases, 

 this melted wax takes the shape of the two arms of the letter X, which 

 we have noticed on the platinum ; in many cases, the jambs are curved, 

 having their convex sides in front. This happens when the heat has 

 spread to a greater extent, and the wax has melted in the whole 

 of the space between them. 



The prism which has this line for its base, and the width of the bar 

 for its height, represents a certain bulk and a certain weight, and if it 

 be admitted that the whole of it has been raised to the temperature of 

 the melted wax, this rise in the temperature must represent a certain 

 quantity of heat, or. by its mechanical equivalent, a certain quantity of 

 interior work, which is directly proved by the experiment. 



By comparing this converted work to the work furnished by the fall 

 of the ram, we find a co-efficient of mechanical return, which is not 

 less than 70 per cent. We do not consider this number definitive ; 

 it depends upon the conductibility of the metal, the solidity of the 

 apparatus used in carrying out the experiment, the cleanliness of the 

 surroundings of the melted surface. But what I wished to point out, 

 Gentlemen, is, that we have come back to M. Joule's first methods, and 

 that our labours on the flow of solid bodies are already bringing us 

 back to the verification of some thermo-dynamical statements. 



s 



