TITLES AND ABSTRACTS OF PAPERS 101 



during or just after solidification, (4) streaked differentiation, (5) successive 

 intrusion, (6) heterogeneous intrusion. Tlie writer would add (7) convection 

 during crystallization. 



Presented in abstract from notes. 



Discussion 



Dr. W. J. Millek: Doctor Grout should be congratulated on this paper. 

 The Adirondack anorthosite shows manj^ bands of varying mineral composi- 

 tion, but not as conspicuously as the Duluth gabbro, and the bands, so far as 

 I have observed them, are extremely variable in dip and strike, with appar- 

 ently no general parallelism to contacts. Possibly the laccolithic structure of 

 the Adirondack anorthosite as opposed to the sill structure of the Duluth 

 gabbro might account for the difference. 



Dr. M. E. Wilson : To one who is constantly encountering the phenomenon 

 of a banded structure in igneous rocks, Doctor Grout's paper is of special 

 interest. There is one way, however, in which I think a banded structure 

 might possibly develop in an igneous rock that Doctor Grout did not mention, 

 and that is by diffusion during consolidation. 



It is obvious that if a magma is homogeneous before consolidation com- 

 mences, the development of individual crystals is in reality differentiation on 

 a small scale. In some magmas this differentiation by crystal growth appar- 

 ently continues until aggregates of crystals have developed and a rock of a 

 very heterogeneous appearance is formed. If a magma in whicli this differen- 

 tiation process was in progress was subject to differential pressure, it seems 

 reasonable to suppose that the segregation would take place linearly, resulting 

 in a banded structure. Many of the igneous Precambrian rocks of the Cana- 

 dian Laurentian highlands are characterized by a minute discontinuous band- 

 ing, the origin of which, it seems to me, can be best explained in this way. 



Eemarks were also made by Prof. James F. Kemp and Dr. M. E. 

 Wilson. 



TWO-PHASE CONVECTION IN IGNEOUS MAGMAS 

 BY FRANK F. GROUT ^ 



(Abstract) 



Convection is indicated by the banding in igneous rocks. It is observed in 

 lava lakes and may be inferred from other rock structures. Convection in a 

 deep magma chamber has usually been attributed to density differences due to 

 temperature of the liquid. Daly also emphasizes the change of density due to 

 a concentration of a gas phase. This paper emphasizes the density difference 

 due to crystal phases. 



The order of magnitude of the forces is estimated. The results of such two- 

 phase convection are outlined. 



Presented in abstract from notes. 



Introduced by W. H. Mmmonf 



