TWO-PHASE CONVECTION IN IGNEOUS MAGMAS 483 
There is no sign whatever of an internal contact between the 
two rocks such as would indicate successive intrusions. Neither is 
there any fluxion structure such as‘a strong extrusive movement 
during crystallization would be expected to produce. It is much 
more likely that the gabbro and granite separated from the same 
liquid, and that this was moved about only slightly, by convection 
or by a late phase of injection, so that the supply of basic minerals 
to the sides was continuous until all had crystallized. 
THE MECHANICS OF CONVECTION IN IGNEOUS MAGMAS 
The nature of convection.—The familiar illustration of convection 
is the local heating of a tank of water, with some suspended matter 
to make its motion visible. The density in different parts: of the 
tank is different enough to start a motion of readjustment, and if 
the difference is maintained by a continuous heat supply and con- 
tinuous cooling elsewhere, a circulation is maintained. As applied 
to magmas the process was suggested by Becker and has been 
widely applied.t. The chief factors which control the rate of 
circulation are the differences in density in different parts of the 
container and the viscosity of the liquid; and both of these factors 
vary with temperature. Water, which is the liquid in most labora- 
tory experiments, shows a very great density change with tempera- 
ture, and its viscosity 1s very low as compared with that of igneous 
magma. Both conditions being exceptionally favorable, it is clear 
that the illustration should not be applied to magmas without some 
estimate of its quantitative importance. 
Two-phase convection.—Besides the well-known changes in 
density in a magma from changes of temperature, there are other 
density effects due to a separation of phases. The separation of 
gases and crystals from lava is a matter of common observation. 
The separation of gases was suggested by Pirsson’ as a possible 
means of deep-seated stirring, and the idea was further developed by 
1G. F. Becker, ‘‘Some Queries on Rock Differentiation,’ Am. Jour. Science, III 
(1897), 21. L. V. Pirsson, ‘“‘The Igneous Rocks of the Highwood Mountains,” U.S. 
Geol. Surv. Bull. 237, pp. 184 and 189. 
2L. V. Pirsson, ‘‘The Petrographic Province of Central Montana,” Am. Jour. 
Sci., XX (1905), 47. 
