STOKES.] 
DATA FOR ARTIFICIAL MIXTURES. 
29 
mixture Avill remain practically constant, and if the value of p be 
obtained by using such an excess of the mixture that this ratio is not 
appreciably altered during the experiment, or if the same amount of 
mixture and of solution be used in every case, it becomes possible to 
obtain concordant values of p and to construct the desired curve. 
The assumption that the ratio of surfaces remains constant during 
grinding is not absolutely true, but practically it is, as the duplicate 
results presented below were each obtained from different samples, 
so that the slightly varying hardness of different specimens and the 
different degree of fineness 1<> which the}' are ground introduce no 
greater errors than those which are inherent in the method itself. In 
each case the materials were broken up so as to pass a 20-mesh but 
to be retained by a 60-mesh sieve, were carefully weighed, mixed, 
and ground so as to avoid all loss, and 1 to 1.02 grams of the mixture 
and 250 cm. 3 of the solution were used in each experiment. 
VII. DATA FOR ARTIFICIAL MIXTURES. 
The following table gives the result for different mixtures, the 
numbers in the second column referring to the above-mentioned 
specimens of pyrite and marcasite, respectively: 
Values of i> for mixtures of pyrite and marcasite. 
Per 
cent 
pyrite. 
10 
20 
40 
60 
M) 
90 
9 
100 
|m.7._- 
m. 10.. 
Im.12 . 
(p. 5; ni. 
p. 5; m. 
Ip. 5; id. 
|p.5; in. 
Ip.5; m. 
| p. 5; in 
lp.5; m. 
p. 5: in. 
p. 5; m. 
jp. 5; in. 
Ip.4; m. 
p. 5; m. 
p. 5; in. 
[p. 5; m. 
\). 4; m. 
p. 5; m. 
p. 4: m. 
p. 4.... 
p.5._._ 
10 
Specimens used. 
16. 
17. 
18. 
15. 
16. 
15. 
15. 
15. 
17. 
16. 
'J 2. 
22. 
29. 
29. 
40. 
40. 
48. 
48. 
52. 
53. 
60. 
60. 
Mean value 
of p. 
17.3(18.0) 
16.0 
15.2 
17.1 
22 3 
29.0 
40.3 
48.9 
52.9 
60.5 
