ON DIMORPHOUS BODIES. 18] 
26. But this conclusion involves several important modifica- 
tions in the received views regarding the atomic weights of other 
substances, elementary and compound. 
It was observed by Mitscherlich that the sulphate of silver 
(AgS) and anhydrous sulphate of soda (NaS) agree in form, 
from which it is inferred that oxide of silver and soda are iso- Equivalent 
morphous. But if so, they are analogous in constitution ; and ae 
if the equivalent of silver be halved, that of sodium must be 
halved also, their formule being respectively Ag,O and Na,O. 
Since, also, potash is isomorphous with soda, and may replace 
it, as in the alums, the rhomboidal nitrates, &c., this oxide also 
must be expressed by K,O. And, on the other hand, gold heing 
isomorphous with silver, the oxide of gold will be Au,O3, which 
agrees also with the results of Dulong and Petit, and with the 
electronegative properties sometimes exhibited by this compound. 
It is unnecessary in this place to dwell on these changes. 
They are indicated by the isodimorphism of the sulphurets of 
copper and silver inserted in the table, but they have not yet 
been incorporated with received knowledge by any of the lead- 
ing chemists of Europe. The establishment of a very few facts 
more will render any further hesitation unnecessary*. 
27. The halving of the atom of potash supplies us with a mode Analogy be- 
of establishing an analogy between the formule for the earthy Saas = 
carbonates and that of the nitrate of potash. If potash be KO i> Carbon- 
and nitric acid, as it is represented by foreign chemists, N,O,, _ and 
Brae itrates. 
then nitre is K,0+N,0; or R’R, or, putting together the posi- 
tive elements R,O,, or 2R,O,. In the carbonates RR we have 
also by putting together the positive elements R,O,, or the for- 
mula for the nitrate of potash is analogous with that for the 
carbonates as a whole, though the expressions for neither of the 
immediate constituents of the two classes of compounds have 
any analogy. 
How far it may hereafter prove true that compounds, as 
such, may be isomorphous and analogous in constitution, while 
their several components disagree both in form and in constitu- 
tion, is at present almost wholly conjectural. I have advanced 
this mode of establishing an analogy between the nitrates and 
carbonates, partly with the view of drawing attention to the 
possible recognition of such a principle as our knowledge ad- 
vances, and partly of illustrating what I have already stated 
(22) as to the special hypothesis necessary in almost every case 
* See London and Edin. Phil. Mag., April 1838. 
