On Crystalline Forms and Chemical Formulae, 407 



Supposing the law to be universally true that like forms in- 

 dicate like formulae, let us inquire what changes in the re- 

 ceived formulae for the several members of each of the above 

 groups it would be necessary to introduce, in order to re- 

 move the discrepancies which they exhibit. 



1. If K O be replaced by N H4 O, it will be necessary to 

 suppose potassium a compound metal, and that it is truly re- 

 presented by a combination R R'4 of two perhaps hitherto 

 unknown elements. 



2. The consequence to be deduced from the mutual re- 

 placement of the sulphurets of copper and of silver has been 

 stated at length in a former paper *. It appears to imply that 

 both sulphurets must be analogous in constitution, and as 

 leading to fewer changes, that the sulphuret of silver is a di- 

 sulphuret Ag2S. 



This again implies that the received atom of silver must be 

 halved, as was already suggested by its specific heat ; and as 

 a consequence, that those of gold, soda, and potash must be 

 halved also. The facts on which this latter consequence rests 

 are stated in the paper above referred to. 



3. The next case, that native sulphur has the same form as 

 bisulphate of potash, seems to indicate only that sulphur, so 

 far from being a simple body, is more complicated in its con- 

 stitution even than potassium, as indicated by the first group. 

 If water may replace potash, then the formula for the bisul- 

 phate becomes R S, which, to bring it into conformity with 

 the supposed law, would require in sulphur a composition 

 represented by R R' + R" R'3. 



4. If from the two formulae in group 4° we take away the 



atom of potash common to both, we have 2S4-H, and 



4 Si + Al respectively. By what hypothesis these two formulae 

 are to be assimilated it is almost needless to inquire. Any 

 method we could adopt would be at least as much opposed to 

 received views as those suggested by the cases already con- 

 sidered. 



5. Can the formulae for nitrate of potash and calc spar be 

 reconciled ? In this Journal, vol. xii. p. 480, will be found 

 a short paper in reply to this question. It is thei'e shown that 

 if the atom of potassium be halved, as our second group sug- 

 gests, and if that of Berzehus for nitrogen be adopted, the 



formula for nitre would become K N, or taking the positive 

 elements together R4 Og. That of calc spar is Ca C, or taking 

 • Lond. and Edinb. Phil. Mag. vol. xii. p. 324. 



