GENESIS OF THE ALKALINE ROCKS II5 
factor in the production of the diversity of igneous rocks.” He 
agrees with ‘“‘the great majority of petrologists”’ that 
the rocks of any area vary among themselves in a systematic manner which 
indicates derivation from a common stock through some systematic process of 
differentiation from that stock. 
The decision is reached that this differentiation is controlled entirely by 
crystallization. The sinking of crystals and the squeezing out of residual 
liquid are considered the all-important instruments of differentiation, and 
experimental evidence is adduced to show that under the action of these pro- 
cesses typical igneous-rock series would be formed from basaltic magma if it 
crystallized (cooled) slowly enough. The characteristic occurrence of basaltic 
magma as regional dikes and as the material of the great fissure eruptions is 
considered evidence of the primary nature of basaltic magma. It is concluded, 
therefore, that most, if not all, igneous rocks have probably been derived from 
basaltic magma, the processes of differentiation that have been pointed out 
above emphasizing the lighter, more salic and alkalic differentiates in the upper 
portions of very large, slowly cooled bodies. 
Definition of ‘‘differentiation.’’—On page 3 of his paper Bowen 
defines differentiation as “‘any process whereby a magma, without 
foreign contamination, forms either a mass of rock that has different 
compositions in different parts or separate masses that differ from 
one another in composition.” In two respects this definition is 
unsatisfactory. 
The word “‘differentiation”’ is advisedly and very generally used 
to mean the actual separation of facies which were once in mutual 
solution or formed parts of the same body of erupted magma. One 
cannot postulate initial homogeneity for every erupted magma, nor 
assume that any heterogeneity a magma possesses has been caused 
by the break-up of an initially homogeneous solution. For example, 
if Suess’s idea, that crust rocks are fused by juvenile gas rising from | 
the deep, hot interior of the earth, should prove correct for any 
eruptive center, initial heterogeneity for the new magma would not — 
be improbable. Each part of it might become still more hetero- 
geneous through partial crystallization or other processes, but the 
original heterogeneity might persist. The rock phases resulting 
from originally different parts of the magma would be different; 
they can be called differentiates only by destroying the useful 
definition of ‘differentiation’ already adopted, expressly or tacitly, 
