Voi^. 6, 1920 
GEOPHYSICS: R. B. SOSMAN 
595 
such complex system. It does seem possible, however, to cover the field 
of actually occurring rocks by partial systems based upon stable mi nerals, 
themselves combinations of two or more oxides. 
Experimental work has progressed far enough to show that the 
fusion and solidification phenomena of the igneous rocks are capable of 
systematic treatment of the kind mentioned above, even when the volatile 
components H2O and CO2 are included. 
Following this fundamental information on the fusion diagrams, we 
must know the densities of the igneous rocks and their magmas at all tem- 
peratures, with their changes of volume during solidification; the textures 
and structures produced by various conditions of solidification ; the latent 
heats concerned in fusion of the magma and in assimilation by it of other 
rocks; and many other physico-chemical data. 
Some of the phenomena of differentiation of silicate rocks are probably 
to be treated as results of simple crystallization backed up by the effects 
of gravity in causing sinking or flotation of crystals. This is the only 
method of differentiation that has been experimentally proven, but the 
separation of two or more liquid phases, and perhaps other phenomena 
also, may take part in this little-known process. 
The phenomena of the movements of igneous magmas in the litho- 
sphere must next be attacked. This involves studies of viscosity and its 
changes with temperature; energy transfers accompanying movement, 
including perhaps the conversion of potential gravitational energy into 
heat energy by movement; differentiation due to movement; and the 
effects of the separation of gaseous constituents in causing differentia- 
tion, movements of magmas, and transfers of energy. 
While the igneous rock is solidifying, processes of metamorphism may 
go on as a result of the passage of its more volatile constituents through 
the igneous rock itself and through its bordering rocks. The rise of tem- 
perature in the inclosing rocks will also initiate reactions and movements 
of material. The chemistry of these processes (alteration, serpentihiza- 
tion, contact metamorphism and replacement, formation of pneumato- 
lytic dikes and veins) will have been included under the chemistry of the 
silicates with volatile components, but the rates of the reactions and the 
transfers of energy accompanying them must be studied as a separate 
problem. The question of the permeability of crystalline substances to 
volatile components may perhaps arise. Studies of the inversion of poly- 
morphic forms, properties of liquid inclusions, and the effects of increased 
temperature on neighboring rocks must be made in order to obtain data 
on the original temperature of the magmas. 
Sedimentary Rocks. — No sooner does the solidified igneous rock find 
itself at or near the earth's surface than it becomes the object of the group 
of processes known as "weathering." This is a subject that has been 
nearly at a standstill since the publication of Merrill's ''Rocks, rock 
