412 
PROFESSOR J. W. MALLET ON THE ATOMIC WEIGHT OF GOLD. 
Experiment. 
Character of 
gold used. 
Gold. 
Silver required to 
precipitate Cl. 
I. 
A, a 
grm. 
7'6U75 
^rm. 
12 4875 
H. 
A, b 
8-4212 
13-8280 
III. 
B 
6-9407 
11-3973 
IV. 
A, c 
3-3682 
5-5286 
V. 
C 
2-8244 
4-6371 
In regal’d to conceivable sources of error connected with this method, it is to be 
observed that, in preparing the original auric chloride solution, if there should be any 
reaction between this o’old salt and the water, leadino: to the formation of traces of 
hydrogen auri-chloride and precipitation of a little auric oxide or hych’oxide, w’hich 
might escipe observation in admixture with the metallic gold left undissolved, the 
tendency would be to lower the atomic weight found for gold. If, by reaction 
between this residual metallic gold and the auric chloride solution, any traces of 
aurous chloride were produced and taken up by the solution of the higher chloride, 
the effect would be to raise the appai’ent value of the atomic weight.'” 
If, in the reaction of the silver solution \ipon that of auric chloride, wdth 
drawal of chlorine should lead to the formation of any traces of aurous chloride, 
precipitated along with the chloride of silver, and not afrerwards decomposed during 
the digestion of the precipitate with the remaining solution, the resulting error would 
also be in the direction of too high an atomic weight. The probability of the last 
supposition is dimimshed by an excess of silver for the whole amount of chlorine 
pre.sent having been added at once. It is not very likely that any one’ of these 
defects actually belongs to the method and affects its results to a sensible extent. Of 
the three I should be more inclined to suspect the possibility of the second than 
either of the two others. 
Second Series of Experiments. 
A neutral solution of auric bromide was prepared by a like process to that used in 
making the auric chloride of the first series : acting upon pure metallic gold with pure 
bromine (prepared with the precautions recommended by Stas), evaporating the 
solution to dryness out of reach of dust, cautious heating of the residue, re-solution 
of auric bromide, and filtration from undissolved metallic gold. 
Two nearly equal portions of the solution were accurately w^eighed off, and treated 
as described above : in one reducino’ the ffold to the metallic state and determinino- its 
O & O 
weight; treating the other with a small excess of silver in solution as nitrate, filtering 
off the precipitate, concentrating the filtrate with the preca.utions already described, 
and determining in it the excess of silver by means of hydrobromic acid. 
* These two remarks apply, of course, also to Kruss’s first series of experiments. 
