416 
PEOFESSOR J. W. MALLET OE THE ATOMIC WEIGHT OF GOLD. 
Expeiiment. 
Character of 
gold used. 
Character of gold salt used. 
Salt ignited. 
Residual gold. 
Loss by ignition, j 
I. 
A, h 
Earlier crop of crystals . 
giun. 
i4-yo72 
grm^ 
— / 3 / o4^ 
grm. ' 
= 7-5318 : 
II. 
A, h 
AJicIcIIg ,, ,, 
15'52d3 
- 7-6831 
== 7-8432 
HI. 
A, h 
La.st „ ,, ,, 
10-4523 
- 5-1712 
= 5-2811 
IV. 
C 
Middle ,, „ „ 
6-5912 
- 3-2603 
= 3-3309 
V. 
C 
Tjixst ,, ,, )) 
5-5744 
— 2-^579 
= 2-8165 
In these experiments the most probable source of error may be fairly taken as 
arising from the presence of traces of methyl-ammonium or dimethyl-ammonium auri- 
chloride with the trimethyl-ammonium salt. I know of no direct evidence that any 
such impurity was present, and the absence of any such evidence in the results from 
the earlier as compared with the later crops of crystals rather tells against the suppo¬ 
sition of its presence, but one cannot feel certain of its entire absence. If present, its 
effect would be to raise the atomic weight obtained for gold. It is also conceivable 
that there may have occurred volatilization of gold to a minute extent as auric chloride, 
in accordance with the observation of Kruss that this salt may be sublimed in small 
quantity at moderate temperatures in a stream of chlorine ; but, there being no such 
stream of chlorine in these experiments, and on the contrary the decomposing action 
of the hydrogen of the triniethylamine salt, this does not seem likely; the effect 
would, of course, be to raise the atomic weight obtained for gold. Another possible 
cause of error might consist in imperfect drying of the gold salt used, but the 
constancy of weight attained on drying renders it unlikely that any other than 
extremely minute error should come of this, though not altogether excluding the 
possibility of its occurrence ; its tendency vrould, of course, be to lower the atomic 
weight obtained. Any trouble from hygroscopic moisture on the surface of the 
porcelain crucible and sand was, I tliink, satisfactorily guarded against by the use of 
a corresponding tare crucible, and by more than one weighing after a near approach 
to the true figures had been obtained, the crucibles having meanwhile been restored 
to the desiccator and kept^ therein for some time. The precautions taken seemed to 
afford sufficient protection against any merely mechanical loss during the ignition. 
Fifth Series of Experiments. 
In these experiments an attempt was made to determine the ratio between the 
weights of metallic gold and metallic silver deposited by the passage of one and the 
same electric current successively through solutions of the two metals. The simplicity 
and accuracy with which the direct weighings may be made seemed to present decided 
advantage, but various difficulties were encountered, and, after the expenditure of a 
very large amount of time and labour upon the method, it cannot be said, on the 
whole, to have satisfied me with its results. 
