MOVEMENTS IN CRYSTALLIZING MAGMA 



FRANK F. GROUT 



University of Minnesota 



The study of the crystallization of silicate melts has led to many 

 important conclusions regarding igneous magmas, several of which 

 are stated in papers recently issued from the Geophysical Labora- 

 tory with the authority of actual experience behind them. Where 

 facts have been discovered, however, concerning a particular 

 hypothesis there has been a tendency to minimize the probability 

 of some other quite independent hypotheses and to carry the con- 

 clusions along into regions not covered by experimental data. In 

 several papers since the publication of Bowen's important outline 

 of the evolution of igneous rocks' the present writer has endeavored 

 to reopen consideration of some hypotheses Hghtly brushed aside 

 in that paper. The arguments have brought forth some new light, 

 but much still remains to be done. 



Liquid magma is known to move as it is intruded, and by con- 

 vection. Crystallizing magma (to accept 'a usage that seems to 

 be approved) is also known to be moved by intrusion and con- 

 vection, and there are molecular movements by diffusion ; in addi- 

 tion there are probably differential movements as crystals settle 

 or float out of magma, and there may possibly be a straining or 

 filter-pressing away from crystals. The magnitude of diffusion 

 is pretty definitely settled by Bowen.^ Little question has arisen 

 over the movements of intrusion, but there is still question as 

 to (i) the activity of convection at depth, (2) the effectiveness of 

 crystal settHng, (3) the mechanics and effects of filter-pressing. 



'N. L. Bowen, "The Later Stages of the Evolution of Igneous Rocks," Jour. 

 Geol., XXIII, Supplement, November-December, 191 5. 



'N. L. Bowen, "Diffusion in Silicate Melts," Bull. Geol. Soc. Amer., 

 XXVII, p. 48. 



2^5 



