TWO-PHASE CONVECTION IN IGNEOUS MAGMAS 487 
added that 10,000 meters of rock would probably increase it no 
more than 30 per cent. This estimate, like the others, may be 
modified roo per cent or more by careful work, but may be taken 
as of the approximate order of magnitude. Applying this addition 
to Bowen’s figures, we have as probable maximum viscosities in 
even figures 5 to 300 in C.G.S: units. 
The effect of new phases remains to be considered. It may be 
assumed that moderate amounts of a gas or liquid phase will have 
little effect on the motion of bodies of magma. The accumulation 
of crystal phases, however, may give a decided difference in results. 
Direct data not being available, it is well to consider analogous 
cases. Curves have been drawn showing the effect of clay added to 
water and dilute water solutions. Though the increase in viscosity 
may be great in some cases, it is shown that a slip with 50 per cent 
solids may have a viscosity less than ro per cent greater than that 
of water. A rough test by the writer, with starch and rock powders 
of about 80 mesh, up to 25 per cent of volume, showed an increased 
bulk viscosity of less than 5 per cent. This would have little effect 
on the maxima above estimated. Bowen estimates that a magma 
may be eruptible even with 50 per cent crystals. The maximum 
viscosities assumed in this paper will therefore be from 5 to 300. 
‘The thermal gradient in magmas.—The variations in temperature 
in different parts of a magma during the cooling process have not 
often been estimated. Estimates of the thermal gradient in a 
magma occupying a chamber may be made from the calculations 
and assumptions of several authors, but they vary from 100° to 
300°C.4 Since it is here argued that convection would occur, let 
us assume that cooling occurs without convection, and calculate 
the forces tending to start such convection. For example, assume 
~C. Doelter, Physikalisch-chemische Mineralogie (1905), p. 110. 
2A. V. Bleininger, U. S. Bureau of Standards Technologic Paper 51 (1915), pp. 
25-30. 
3N. L. Bowen, “Later Stages of Evolution of Igneous Rocks,” Jour. Geol., 
Supplement, December, 1015, p. 31. 
4R. A. Daly, Igneous Rocks and Their Origin, pp. 224 and 258; A. Harker, op. 
cit., p. 316; A. C. Lane, ‘‘Coarseness of Igneous Rocks,’ Amer. Geol., XX XV (1905), 
71; Ingersoll and Zobell, Mathematical Theory of Heat Conduction, etc. Ginn & Co., 
1913. 
