VOLUME XXIx NUMBER 4 
THE 
BewORNAL OF GEOLOGY 
MAY-FUNE r921 
DIFFUSION IN SILICATE MELTS 
N. L. BOWEN 
Geophysical Laboratory, Carnegie Institution of Washington 
INTRODUCTION 
Many well-known features of igneous rocks are clearly the 
result of diffusion. The growth of crystals is accomplished largely 
by the diffusion of material to crystalline nuclei. Inclusions often 
exhibit solution at their borders and some diffusion of their material 
into the surrounding magma. ‘These phenomena involve diffusion 
through short distances and there is no theoretical objection to the 
acceptance of diffusion at the required rate. At the same time it 
has often been inferred that large quantities of matter have dif- 
fused through considerable distances. ‘This has been done for the 
purpose of explaining various associations of igneous rocks and 
especially the formation of basic border phases. Since Becker’s 
destructive criticism of this view’ many geologists, but not all, 
have been less willing to assign to diffusion any important rdéle 
in the production of such features. 
Becker’s objection was based on the extreme slowness of 
diffusion in all cases where its rate had been determined. The 
principal data then available referred to the diffusion of salts in 
aqueous solution, and it was on the basis of reasonable deduction 
t American Journal of Science (4) Vol. III (1897), p. 27- 
205 
