﻿ATOMIC HEAT TO CRYSTALLINE FORM. 117 



synthesis of scientific facts. In these investigations it must, for the present, be kept 

 constantly in view, that the numbers obtained by experiment for the specific heat of 

 different bodies, do not accurately express the amount of caloric combined with the 

 ponderable matter of those bodies. The specific heat is only a portion of the whole 

 quantity of combined heat; and, as Regnault has shown in his second memoir,* the 

 obstacles to be overcome in obtaining the specific heat of any substance, with perfect 

 accuracy, are exceedingly great, if not, indeed, insurmountable with our present 

 appliances. 



From the isomorphism of their double chlorides, Dr. Clark is inclined to associate 

 the. tin and platinum classes. The relation of these two series in heat and volume 

 lends considerable support to their presumed affinity. According to Breithaupt, CdS 

 is isomorphous with the alloy IrOs. The atomic volume of either iridium or osmium 

 is to that of CdS as 1 : 1 i. Palladium, although not generally arranged in this series, 

 appears here to find its natural place. 



Tungsten and molybdenum correspond in heat and volume. The tungstates and 

 molybdates are isomorphous, as their agreement in atomic volume sufficiently indicates. 

 I have not been able to find upon record the atomic heat of these salts. Their sup- 

 posed isomorphism is confirmed by the following comparison : — 



At. Heat. At. Volume. 



W0 3 .... 9.496 . . . . 22.580 

 M0O3 .... 9.533 . . . 20.809 



The bisulphurets of tin and molybdenum, which present some resemblance in their 

 physical constitution, have atomic heats but little different. The bisulphuret of iron 

 differs very much from these in its characteristics, and we find that its atomic heat 

 is also very different. (Regnault.) 



In the metallic state gold and silver are isomorphous. Their relationship in heat 

 and volume is very obvious. I have followed Gmelin in adopting 199 as the atomic 

 weight of gold. If we assign it 99.6, as do some of the chemists, and halve the atomic 

 weight of silver, for the cogent reasons advanced by Regnault, the above relation 

 still holds good. Or, if we reject the suggestion of Regnault, it will be found that in 

 heat and volume silver is the exact multiple by two of gold ; so that in either case, 

 the simplicity of relation (which in the present state of our knowledge of these 

 subjects is all that we can contend for) is still preserved. Sodium allies itself with 

 silver, as shown by the resemblance in atomic volume of their carbonates, sulphates, 

 nitrates and chlorides. In all this series, as far as the specific heats have been 

 obtained, the same correspondence in heat and volume is evident. 



The atomic volume of sodium, as given in the table, is about one-half that of potas- 

 sium. These volumes would be equalized by doubling the atomic weight of sodium ; 

 and it is interesting here to observe that this alteration has been proposed by Dr. 

 Clarke from theoretical considerations of a different character. The simplicity of 

 relation is therefore evident. 



* Annales de Chim. et de Phys., 3<ne S er., t. 1, p. 200. 



