100 
ON THE DOUBLE SALTS OF IRON. 
the examination did not corroborate such an idea : never- 
theless, both a tercitrate and quadricitrate may exist, and 
the latter supposition gains support from the probability of 
acetic acid combining in the same proportion. 
Sesquioxide of iron is not precipitated from its combina- 
tion with citric and tartaric acids, by alkalies, alkaloids, 
earthy and some metallic oxides : according to Professor 
Brande, malic, mucic, kinic, and pectic acids lead to the 
same result. It has been found that^double salts are formed, 
the precise constitution of which chemists are by no means 
agreed on : my own observations induce me to regard a bi- 
nary series, the existence of which I will endeavor to de- 
monstrate. 
I took bicitrate of ammonia, and to its boiling solution 
added hydrated oxide of iron so long as a greenish-yellow 
color was retained, or so long as a distinct red absented 
itself : the solution must be boiled a few moments after 
each addition of oxide, or the effect will not be observed : 
when the process is complete, we have a combination of 
one equivalent of bicitrate of iron with two equivalents of 
citrate of ammonia. I obtained from one specimen so made, 
very nearly twenty-two grains of oxide per cent., and from 
another twenty-four grains : it dries in scales of a golden 
color. 
Again, I took the precipitated oxide derived from four 
parts and a half of commercial sesquioxide, and boiled it 
with five parts and a half of crystallized citric acid, and 
after continuing ebullition a short time, it was supersaturat- 
ed with caustic ammonia ; it was then filtered, and cooled 
quickly, and I had a solution most difficult to evaporate 
without, decomposition, but which, when dry, was perfectly 
soluble : one hundred grains of this salt afforded me nearly 
thirty-six grains of oxide of iron. I am sorry to say, another 
gentleman, to whom I gave a portion of this salt, only ob- 
tained thirty -four grains: pardon my attaching importance 
to this circumstance. It was my request that before exami- 
