PROFESSOR J. W. MALLET ON THE ATOMIC WEIGHT OF GOLD. 
413 
Pteducing the amounts of silver actually used to the corresponding quantities for 
portions of auric bromide solution exactly equal to those from v/hich in each case the 
gold was obtained, the results in six experiments stood as follows ;— 
Experiment. 
Char mte*' of 
gold used. 
Gold. ' 
Silver required to 
precipitate Br. 
sirni. 
grm. 
T. 
A, h 
8-2;345 
13-5 J 49 
H. 
A, c 
7-6901 
1-2-6-251 
Ilf. 
B 
10-5233 
17-2666 
IV. 
A, a 
2-7498 
4-5141 
V. 
C 
3-5620 
5-8471 
VI. 
A, h 
3-9081 
6-4129 
In these experiments the sources of constant errors which suggest themselves as 
possible are essentially similar to those for the first series; but, if any such really exist, 
there is, of course, the likelihood of some difference being introduced by the substi¬ 
tution of bromine for chlorine. Hence the desirability of multiplying experiments in 
this modified form. 
Third Series of Experiments. 
For these experiments potassium auri-bromide was prepared with great care from 
an excess of metallic gold treated with bromine and potassium bromide, purified in 
accordance Vvdth Stas’s suggestions, and the double salt five times recrystallized The 
last crystallization was conducted fractionally, in closed vessels, with special care to 
exclude dust, by gradual but pretty rapid cooling with agitation, and the earlier and 
later portions separated out were kept apart in after use. 
For each atomic weight determination an unweighed quantity of this potassium 
auri-bromide was dissolved in water, the solution rendered uniform by agitation, and 
divided into two nearly equal parts, which were severally weighed with accuracy, and 
in one the gold reduced to metal as in the experiments of the first and second series, 
and in the other the total bromine precipitated by silver solution as before, the com- 
})arison being made once more between the weight of the gold and that of the silver 
equivalent to the bromine (m this case representing 4 atoms) existing in the double 
bromide. 
Again stating the quantities of silver corresponding to portions of the auri-bromide 
solution exactly equal to those used in determining the gold, the following were the 
results obtained : — 
