236 
SELECTED ARTICLES. 
to say that the most probable rational formula is that here 
subjoined: 
IK + 3 (I + Cin 2 .) 
From the analysis which I first performed, and of which I 
gave a brief account in the Chemical Section at the Liverpool 
Meeting of the British Association for the Advancement of 
Science, the formula deduced was 
IK +2 (I 2 + Cin 3 ,) 
which differs from the preceding merely in containing one 
more atom of iodine. 
This compound appears interesting under many points of 
view. In the first place, it is one of considerable complexity, 
is decomposed with an extreme facility, and is nevertheless 
perfectly definite in its composition, and even beautifully 
crystallized. 
In the second place, it is a kind of double salt, composed of 
two haloid salts, in one of which the oil performs the very 
unusual function of an electro-positive or basic metal, — a cir- 
cumstance the more singular, as Dumas has shown that it 
unites also to the muriatic and nitric acids, forming with them 
binary compounds, the latter of which very readily crystal- 
lizes. The oil, in fact, thus appears to act the part of a metal 
as well as of an oxide. 
Lastly, I may observe that the method by which our com- 
pound was first accidentally formed, and is still best made, 
presents an instance of incompatibility which had not been 
previously suspected, and will, no doubt, suggest to chemists 
experiments which will eventuate in the production of a series 
of similar substances. In reference, however, to this latter 
point, I should add that Mr. Moore has applied to the other 
aromatic waters the very process which succeeds with cinna- 
mon water, but without obtaining a trace of any new product. 
It is possible, however, that new results might be obtained 
by substituting other metals for the potassium, and replacing 
