RELATIONS TO CLASSIFICATION AND MINERAL TYPES. 59 
different types of natural occurrence: the phyllid and the stone type, respectively. May 
we not therefore assume that there may also be (cryptomorphically) a metallic carbon, a 
thionoid carbon, &e., as well as a graphitic sulphur, and so forth; and that in graphite 
we haye an atom of metallic carbon combined with two atoms of graphitic carbon, repre- 
senting the S’ (in its phyllid or graphitic condition) of molybdenite? Or it might be 
assumed that both components of the latter mineral have passed into their graphitic or 
phyllid form. At any rate, explain the fact as we may, there must be some definite cause 
for the very striking physical resemblances of the two minerals; and this cause may be 
found, I submit, in the assumption of cryptomorphism, or erypto-allotropism, as induced 
by contact in elementary bodies, generally. 
The preceding remarks, and their natural deductions, may be summed up briefly as 
follows : : 
1. Elementary bodies are capable of occurring—as long known—either visibly or 
cryptomorphically in very different physical conditions. 
2. These conditions are principally, the metallic condition, the phyllid condition, the 
thionoid condition, and the stony condition, as defined above: with others of an inter- 
mediate or subordinate character. 
3. The contaet of two elementary bodies in combination, induces generally if not 
invariably a changed condition in one, or in all, of the combined bodies—the changed bodies 
becoming totally distinct from the same bodies in their visible or normal forms of occur- 
rence. 
4. Hence, whilst minerals may be classed for technical ends in accordance with the 
metal, or the electro-negative principle, present in them, or otherwise in accordance with 
their stochiometrical relations, classifications of this kind are utterly wanting in all that 
constitutes a natural system, and lead in many cases to strikingly incongruous groupings. 
5. Thus—although commonly so placed—graphite, as a mineral, cannot properly be 
placed with the diamond in one and the same class; nor native iron with native lead; 
nor iron pyrites with molybdenite; nor pyrrhotine (magnetic pyrites) with cinnabar ; 
nor periclase with litharge (simply because both are oxides of the formula MO); nor 
corundum with bismuth ochre and valentinite ; nor quartz with pyrolusite or with cer- 
vantite ; nor manganite with sassoline (B° O°’); nor magnetic-iron ore with minium ! 
These and other, from a mineralogical point of view, equally astonishing collocations are 
to be found in the last edition of a well-known German treatise of wide reputation, and in 
other modern works which profess to be exponents of mineral affinities. To construct a 
classification of this sort, founded upon a single character, in itself more or less conventional 
or unsettled, it is clear that one need not have the slightest acquaintance with minerals, 
beyond a knowledge of the bare results of their analysis. If minerals occurred only in the 
condition of fine powder of uniform character, they might be legitimately arranged in 
reference to composition pure and simple, or to their atomic relations; but minerals are 
something more than abstract chemical bodies. They are evidently distinct individualiza- 
tions, endowed with independent characters of their own : and these characters cannot be 
ignored in a truly scientific or natural classification. 
