GENESIS OF THE ALKALINE ROCKS sat) 
to cause gravitative adjustment among the liquid parts of an 
originally heterogeneous magma. ‘The parts richest in gas are 
likely to rise roofward, displacing gas-poor parts of otherwise 
similar composition. A xenolith of gneiss, melted by, but not 
diffused into, a moderately viscous gabbroid magma, would rise 
rapidly and segregate with other melted bodies of the same kind. 
The most fusible part of the disintegrating xenolith would rise or 
sink independently. Even if diffused, the more acid foreign mate- 
rial locally lowers the density of the magma, for diffusion is a slow 
process. On the other hand, the local masses of acidified gabbro 
will rise quickly if the viscosity of the main body of gabbro is low. 
Differentiation of the units described in classes 1, 5, and 6 may 
take place without any crystallization whatever; that is, all the 
phases concerned are liquid. Yet in none of the cases is the prin- 
ciple of fluid immiscibility necessarily concerned. The gas, juve- 
nile or resurgent, like the xenolithic gneiss, may be perfectly 
miscible with the original magma; nevertheless gravitative adjust- 
ment is compelled long before homogeneity could be produced by 
the diffusion of foreign matter. 
Whether true liquid immiscibility is an additional, perhaps very 
important, factor in magmatic differentiation is uncertain. Yet 
the repeated dogma that this question must be answered in the 
negative for natural silicate solutions is not warranted. The 
proper answer awaits the time when the influences of undercooling 
(by pressure, etc.) and volatile agents, as well as other unknown 
conditions, are better understood than now. 
Bowen has done good service in confirming the view that the 
diversity of igneous rocks may be partly due to the sinking (or 
rising) of early-formed crystals in magma, but he has carried the 
principle farther than it is safe to carry it in the light of present 
knowledge and in the dark of present ignorance. His definition 
of differentiation is subjective, since that process is assumed to 
affect only purely juvenile magma “without foreign contamina- 
tion.”’ With respect to the biggest problem in petrogeny he thus 
takes a position which a host of field facts renders untenable. 
Certain observers ascribe to magma the power to dissolve com- 
pletely large percentages of country rock. Others, more cautious 
