312 The American Geologist. November, 1897 
every other paragenesis of this mineral, partial peroxidation 
has supervened for the production of ferric oxide in atomic 
ratio with ferrous oxide. Saturation of ferrous oxide with 
ferric oxide for this evolution is indeed all that has arrested 
peroxidation of the lower oxide short of completeness wher- 
ever original silicates have been split up by oxidation of this 
unstable base under Aveathering influences. Hydro-chemical 
as well as thermo-dynamical agencies in eruptive paragenesis 
of magnetite have doubtless been far more active than in a 
metamorphic paragenesis. Physical differentiation or mag- 
matic concentration, of magnetite as such into concrete ore 
bodies from molten eruptives, as urged by Vogt, is scarcely 
conceivable without superventi(»n of further physical or molec- 
ular concentration analogous to what takes place both in 
metamorphic and eruptive aggregates. No basic rocks even 
among the most modern can ever be assumed to retain their 
original condition or inter-molecular form. 
There seems much however to support the theory that 
differentiation in molten eruptives may summarily take place 
in the order of specific gravit}^ toward the borders of a magma 
basin, or, on the Soret principle, toward the marginal parts of 
dykes or other intrusions, with the effect of a determination, 
not necessarily in stochiometric proportions, of material 
richest in ferrous oxide to such a position relative to the 
mass. Hence of course the greater basicity of silicates and 
also the greater the subsequent evolution of magnetite in 
nether parts of intrusive masses, sometimes so as to afford 
the locus of ore bodies through slow hydro-chemical permu- 
tations. That concentration of magnetite has taken place in 
any given eruptive magma except as a product of secondary 
evolution seems to me extremely doubtful, as in the case of 
all other bodies of iron ore closely related to basic rocks — 
notabl}"- metamorphic rocks. The subject of physical differen- 
tiation of igneous magma antecedent to development of 
crystalline types of igneous rocks has been presented by Prof. 
Iddings in a manner which leaves little room at present for 
further citations of results of studies given to the same sub- 
ject by Brogger, Vogt and others.* 
*Iddings. Bull. Phil. Soc. Washington xii, 1892, pp. 89, 214: Vogt. 
Zeitsch. fur. prak. Geol. i, 1893, pp. 4, 125, 257; iii, 145, .367, 444, 465. 
Note. Since the presentation of this paper an exact discussion of 
this hypothesis has been given by Becker. (Am. Jour, of So. iii, 1897, 
21.) 
