stokes.] INFLUENCE OF IMPURITIES. S3 
in order to render the precipitation complete and to obtain a correct 
value for e. 
An experiment made with the usual precautions, with carefully 
selected arsenopyrite, gave — 
<i, 39.50 
b, 56. 90 
c, 45. 84 
On substituting these permanganate values we get 
p = 49. 8 
The actual sulphur oxidized, as calculated from these figures, assum- 
ing that all the arsenic is oxidized to As 2 3 , is 10.3 per cent, and in 
fact a notable sublimation of sulphur was observed. The data do 
not indicate whether any arsenic is liberated, or whether orpiment 
or realgar is formed, but no evidence of these could be obtained. 
More elaborate experiments would be required fco decide whether 
under certain conditions they can be formed by the action of ferric 
salts on arsenopyrite. 
A carefully prepared mixture of pyrite and arsenopyrite (2.17 per 
cent), containing 1 per cent arsenic, was oxidized with the following 
result: 
</, :5S.33 
Z>, 4:J.17 
c, 42.::! 
The value thus deduced for j), namely, 5(5.7, corresponds to that of 
a pyrite containing 2 per cent marcasite. As few pyrites contain as 
much as 1 per cent arsenic, it appears that in general the presence of 
arsenic will not interfere greatly with the marcasite determination. 
CHALCOCITE AND BORNITE. 
Chalcocite by itself gives an infinite value for p. The influence of 
these minerals on pyrite is state* 1 below. Both are very easily attacked 
by cold ferric solution. 
CHALCOPYRITE. 
As pointed out in the introduction, it is well known that chalcopy- 
rite is more easily attacked by ferric solutions than is pyrite. An 
experiment with ciystallized chalcopyrite made with the usual pre- 
cautions gave 
a, 37.44 
b, 44.01 
c, 44.11 
whence 
P=m.i) 
From these figures and the determination of the dissolved copper 
it appears that not more than 2 per cent of the sulphur is oxidized, 
and that more copper than iron is dissolved (0.7 atom Fe to 1 atom 
