on the Haloid Compounds of Mercury. 425 
Having, by careful repetition of the above method, obtained a product such as I 
might consider pure, I entered upon its analysis. I think it advisable, however, before 
commencing the details of my own experiments, or noticing the results to which they 
have led, to bring forward a tabular statement of the analytical results of other writers, 
in order to show in what obscurity the subject was involved. ‘The results are taken, 
divested of the theories to which they have respectively conducted their authors, and 
are exhibited in the real quantities of chlorine, quicksilver, and ammonia, which were 
obtained as the products of the experimental trials. 
There were obtained from 100 parts : 
Authors. Mercury. Chlorine. Ammonia. 
Fourcroy,* 74,1 13,2 6,03 
Hennell,t 74:,24 13,14 6,31 
Mitcherlich,t 76,37 13,82 vial! 
Guibourt,§ feck) 13,3il 4,45 
Soubeiran, || 82,1 7,8 5,0 
A glance is sufficient to show the impossibility of deducing from such a chaos any 
general expression for the composition of white precipitate. In attempting to find 
some principle by which it could be set in order, I rested my probabilities of success 
on the number of trials I would make ; for, conscious that the chance of incorrectness in 
any one analysis should be still greater with me than with any one of those chemists 
whose ill success I have exhibited above, I trusted to diminishing the errors by taking 
averages of numerous results, and obtaining the results themselves by processes dif- 
fering in principle, so as to render it very unlikely that any one error could pervade 
all. I shall therefore describe in order the results of each different method of 
analysis. 
A.—When a solution of corrosive sublimate is precipitated by ammonia, the 
liquor contains no mercury, but much chlorine. All the mercury of the sublimate 
is contained in the white precipitate, and a portion of the chlorine is removed, and 
remains in the solution as sal-ammoniac. It is evident that on this principle may be 
founded a method of determining the amount of mercury and chlorine in white pre- 
cipitate ; and this was accordingly that first made use of. 
One hundred grains of corrosive sublimate were dissolved in cold water, the solution 
decomposed by a slight excess of ammonia, and the precipitate thrown on a weighed 
filter, and washed with cold water. The precipitate was carefully dried and weighed ; 
the liquor added to the washings was acidulated by nitric acid, and precipitated by solu- 
* By result given by Leopold Gmelin, Handbuch. 
+ Quarterly Journal of Science. 
$ Poggendorff, vol. Ixxxv. 
§ Result given by Thenard, Traité Elementaire. 
|| Journal de Pharmacie, vol. xii. 
