28 ON PYRITE AND MARCASITE. [bull. 186. 
density, exactly proportionate in most cases to the amount of each constituent, 1 
and by a similar variation in other physical properties; e. g., hardness, fracture, 
resistance to decomposition, and even in color, in the case of the paramorphs of 
marcasitic pyrite. 
As a paramorph consists simply of minute crystals of the one sub- 
stance massed together in a form which characterizes the other, it is 
obvious that the oxidation coefficient must . correspond to that sub- 
stance of which the specimen really consists, not to that which it imi- 
tates, and such a curve would indicate the proportions in the case of 
a mixed crystal. It would further enable us to follow the possible 
artificial transformation of one into the other, and would serve as an 
aid in synthetic work. In such work it is frequently difficult to obtain 
products sufficiently well characterized to admit of positive identifi- 
cation in the absence of optical properties and well-developed crys- 
tallization. Such a curve would render us independent of these. 
It is not possible to calculate the curve for mixtures of pyrite and 
marcasite from the oxidation coefficients of the two minerals. The 
oxidation coefficient is influenced by the concentration of the ferric 
and ferrous salts and the sulphuric acid. With the pure mineral 
these vary in a perf ectly constant manner in the case of each sulphide, 
but when each mineral is decomposing in the presence of the prod- 
ucts formed by the other, nothing short of an elaborate investigation 
could give us the data for calculating the effect in any case. It is 
therefore necessary to construct the curve empirically, from data 
obtained from a sufficient number of artificial mixtures. 
CONDITIONS UNDER WHICH THE OXIDATION COEFFICIENTS OF 
MIXTURES MAY BE DETERMINED. 
We have seen that the oxidation coefficients of pyrite and marca- 
site are independent of the relative amounts of solvent and solute, 
and of the degree of comminution, and are constant for a given tem- 
perature and concentration provided the reduction be complete. The 
case of a mixture of the two is more complicated. Here each com- 
ponent deprives the other of a portion of the ferric salt, this effect 
being greater the greater the surface, or, what is the same, the finer 
the powder. An artificial mixture of given composition will there- 
fore show a varying oxidation coefficient depending on the relative 
degree of fineness of the two components. Moreover, since a given 
volume of the standard ferric solution decomposes about 2. 20 times 
as much marcasite as pyrite, the ratio between the surfaces is con- 
stantly changing and it is no longer a matter of indifference whether 
a large or a small amount of material be acted on by a given volume 
of solution. 
If, however, the minerals be ground together and if the grinding be 
tolerably thorough, then, considering their approximately equal 
hardness, it may be assumed that the ratio of the surfaces for a given 
Compare table on p. 13. 
