﻿116 



MEIGS ON THE RELATION OF 



rank with sulphur, so the resemblance in their chemical relations between HO and 

 CuO is supposed to connect hydrogen with the group in question. It is curious to 

 observe that the atomic heat and volume of HO tend to confirm the notion of its 

 connection with the magnesian class. 



Lime (CaO) probably occupies an intermediate position between this and the barium 

 series, blending them gradually together. This is very well shown in the table of 

 isomorphous salts, in the two groups of carbonates, where carbonate of lime as calc- 

 spar, assimilates, on the one hand, with the corresponding carbonates of magnesia, 

 iron, the mineral dolomite, &c, and on the other, as arragonite, identifies itself with 

 the carbonates of strontia, baryta and lead. The protoxides of barium, strontium, 

 calcium and lead, are nearly related in heat and volume. But the similarity of the 

 barium class to the magnesian is shown through their chlorides, where the corres- 

 pondence in atomic heat and volume is quite remarkable. 



Salts. 



Atomic 



Atomic 



Heat. 



Volume. 



BaCl 



9.31528 



27.513 



SrCl 



9.52006 



28.357 



CaCl 



9.17878 



26.244 



MgCl 



9.36026 



... 



MnCl 



8.98065 





ZnCl 



9.20577 



... 



SnCl 



9.59198 





PbCl 



9.24427 



25.217 



HsjCl 



9.42415 



25.714 



In the next two groups the numbers for atomic heat are the same throughout; as, 

 with one exception, are those for atomic volume. This is an exception in appearance 

 only, and not in reality, for its volume-number is just about double that of the others, 

 preserving, therefore, a simple multiple relation. This view finds confirmation in the 

 isomorphism of the oxides of tin and titanium, which compounds, as the table shows, 

 are sufficiently identified in heat and volume. Indeed the coincidence is still more 

 positive than at first sight appears. It will be seen that the specific heat obtained 

 by Avogadro for Sn0 2 differs so much from the other three observations, that it may 

 very properly be omitted in calculating the mean specific heat of this body. This 

 procedure will give 6.9075 for the atomic heat of the oxide of tin. So the number 

 furnished by Hermann for the specific heat of Ti0 2 may, with equal propriety, be 

 rejected in ascertaining the mean specific heat of titanic acid. We thus obtain 6.9417 

 for the atomic heat of this latter substance. These two numbers may be considered 

 as identical. I may here remark that a similar analysis of the whole table would 

 lead to results still more striking than those indicated. I have preferred, however, 

 to give all the recorded specific heats, with the mean for each elementary and com- 

 pound substance, from which to calculate the atomic heat, because in this manner is 

 avoided the confusion of so many analogous numbers for any body, and still more 

 because I regard the results obtained in the table as approximations merely to a great 

 general truth or law in nature, which yet remains to be demonstrated by a logical 



