ON THE DOUBLE SALTS OP IRON. 
103 
Take a bisalt of any third constituent (bicitrate of am- 
monia for example) made by adding a known quantity of 
citric acid to a similar quantity exactly neutralized with 
ammonia. This solution in its turn is to be saturated with 
hydrated oxide, which is accomplished by mixing the oxide 
gradually so long as it continues to be dissolved, avoiding 
any great excess. This formula admits of very general ap- 
plication by simply changing the acid, or the third constitu- 
ent, but there are exceptions. 
Dr. A. T. Thomson observes on the ammoniacal citrate, 
that it disagrees with some constitutions more than its con- 
gener, the potassio-tartrate. Guided by this opinion, I 
would direct attention to the compound of tartaric acid, 
iron, and potash, obtained by dissolving the hydrated bitar- 
trate in a solution of the caustic alkali ; if they be mixed 
sulphate to be similarly treated with caustic ammonia. My own plan 
in detail differs from both of these. I make a solution of the sesquichlo- 
ride according to the ^directions of the London College, adding nitric 
acid to insure the complete peroxidation of the iron, but instead of de- 
composing cold, diluted, and with a caustic alkali, I mix the solutions 
warm, somewhat concentrated, and use a carbonated alkali, and for these 
reasons : a moderate degree of heat is favorable to chemical action, and 
when conjointly with this, double elective affinities are in operation, 
decomposition is more complete, and the liability to form a basic salt is 
considerably diminished. Again, if the solution is very weak, the pre- 
cipitate will be of so buoyant a nature that the process of ablution will 
be tedious. As soon, however, as the oxide is precipitated, longer re- 
tention of heat is absolutely mischievous, and large quantities of cold 
water must be added without delay, and the washing continued until the 
supernatant liquor is tasteless ; the remaining water being drained off, 
the oxide is ready. To this it has been objected, that carbonic acid has 
no affinity for sesquioxide of iron. 1 apprehend that affinity does exist, 
although ordinarily it does not influence to combination ; how else shall 
we explain the fact that solutions of sesquicarbonate of ammonia and bi- 
carbonate of potash dissolve, under pressure of carbonic acid, sesquioxide 
of iron ? That an affinity exists, I think is manifest ; and that it pre- 
vents failures in making the hydrated oxide, I am inclined to believe : 
at all events, I have not formed a basic salt since 1 have adopted the 
plan of decomposing with carbonate of ammonia. 
