40t 
GLASGOW CHEMISTS AND DEUGGISTs’ ASSOCIATION. 
portion of the bismuth is thus retained in solution as ternitrate; while, if added to 
excess, the potash at once combines with the citrate of bismuth to form soluble potassio- 
citrate, v/hich is carried off in washing the precipitate. This process, therefore, requires 
the finest chemicals and nicest manipulation to work it successfully. The reason for 
adding the potash and citric acid separately, instead of combined, as Mr. Bartlett directs, 
is that the chance of getting a soluble precipitate is greater by Mr. Blount’s than by 
Mr. Bartlett’s method. 
The chief'objections to the citrate of bismuth process are—1st, That there is a con¬ 
siderable loss, on account of the solubility of the citrate, in the nitrate of potash solution 
from which it is precipitated, necessitating an analysis before the solution can be made 
of a definite strength ; and, 2nd, That the precipitate is often not wholly soluble, either 
in ammonia or citrate of ammonia. The quantity dissolved in the washing, and there¬ 
fore practically lost, is equal to 5 or 6 per cent, oxide ; but the loss on account of 
insolubility in the solution of ammonia sometimes exceeds this, and makes in all a loss 
varying from 7 to 12 per cent. Mr. Ebert puts the total loss down at 7‘G per cent.; 
but this must be the minimum, as, on one occasion, I found the loss by this process to 
be as high as 22 per cent. 
The solubility of citrate of bismuth in the solution from which it is precipitated is 
much influenced by the proportion of nitric acid present. If there be an excess, but a 
small quantity of citrate will be thrown down; if too little, part of the carbonate re¬ 
mains unaltered, and the maximum proportion of citrate is precipitated. The quantity 
of nitric acid given by Mr. Ebert is not sufficient to dissolve the carbonate of bismuth, 
and part is therefore carried through the whole process unaltered, without being taken 
up in the solution of ammonia, and a loss is thus sustained. 
By Mr. Tichborne’s method of dissolving the oxide in citrate of ammonia, the loss is 
much greater than by the above process. The insoluble oxide might again be dissolved 
in nitric acid, precipitated, and added to another portion of the citrate ; but this gives 
much unnecessary labour, and consequently considerable loss in manipulation. On this 
subject Mr. Ebert says—“Teroxide of bismuth, though recently precipitated, is but 
sparingly soluble in citrate of ammonia even at the boiling temperature: the presence 
of a portion of nitrate of ammonia, through insufficient washing of the teroxide, will 
effect a solution.” This is the case if the quantity of nitrate is great; but, if the 
oxide is but slightly w^ashed, the nitrate of ammonia retained apparently in no way 
assists solution. 
Having observed, in following the oxide process of Mr. Tichborne, that the precipitate 
had sometimes a lustrous silky or pearly appearance, and that after being boiled in 
citrate of ammonia, the insolublp part was wholly composed of these lustrous particles, 
I inferred that oxide of bismuth, under certain circumstances, assumed the crystalline 
form, and that in this state it in a great measure resisted the action of citrate of am¬ 
monia. Subsequent observation has borne out this theory. Under a powerful micro¬ 
scope the lustrous oxide appears to be composed of partially dissolved crystals, with 
scaly and amorphous particles interspersed—the proportion crystalline and amorphous 
varying very much in samples from different precipitations. When the crystalline 
or scaly particles predominate much over the amorphous, the oxide is found to be 
very insoluble ; while, if but few crystals are perceptible, very little is left undissolved. 
Investigating the cause of this condition of oxide of bismuth, I found that, as a 
rule, a crystalline oxide was thrown down if strong solution of ammonia, or an ex¬ 
ceedingly dilute solution, was employed in precipitation; and if sufficient ammonia 
was not added at once, and the oxide suffered to remain for some time in con¬ 
tact with the acid liquor, the same effect was produced. Christison has apparently 
noticed this appearance in the trisnitrate. Making remarks on its preparation, he says— 
‘‘ We have to guard against the pulverulent precipitate assuming the crystalline form, 
which is done by throAving the nitric acid solution into a large quantity of water, and 
not alloAving the precipitate to remain long in contact with the acidulated water. By 
attending to this a trisnitrate, in an exceedingly minute state of division, is obtained.” 
All attempts at producing an oxide free from crystals, by A^arying the strength or quan¬ 
tity of ammonia or nitric acid, I found to be fruitless: the same effect being produced 
in apparently quite opposite circumstances ; a particular process producing a precipitate 
at one time nearly amorphous, and at another highly lustrous and crystalline. 
Having observed that citrate of ammonia imsent in the nitric acid solution deter- 
