494 FRANK F. GROUT 
The relative slowness of motion of sinking crystals is evident 
from a comparison of the tables. In a sill 1,000 feet thick a large. 
crystal of olivine might settle in a day or two, though it is not 
probable; small ones would require a few weeks. Convection 
might carry them down in half an hour, or much less if the viscosity 
was not at a maximum. 
TABLE III 
EstIMATED RATE OF FALL OF SETTLING CRYSTALS OF 4 MM. DIAMETER IN 
METERS PER Hour 
Viscosity Mineral Rae ot | Viscosity Mineral Bare 
5 Magnetite 132 300 Magnetite 2 ose 
5 Olivine 25 | 300 Olivine 0.5 
5 Augite ; 25 | 300 Augite l ©eS 
5 Plagioclase (rising) | 13 | 300 Plagioclase (rising) 0.2 
THE PROCESS OF SOLIDIFICATION WITH CONVECTION 
Form and size and composition of magma.—Small bodies of 
magma are usually cooled to crystallization too rapidly to permit 
much circulation. However, the actual limit in size is not to be 
stated, as the fluidity and duration of crystallization may in case of 
abundant mineralizers allow an effective convection in a four-foot 
dike. Thin, tabular masses are unfavorable to any general circu- 
lation after intrusion is complete, but no special form or position 
of the mass seems to prevent all circulation. If the mass is nearly 
horizontal, settling crystals are more effective than circulation. 
It has been suggested also that thick, strikingly dome-shaped 
laccoliths assume their form because of unusual viscosity." In so 
far as this is so they would be unfavorable to convection. The 
viscosity of a magma during crystallization depends largely on its 
composition. Data are not abundant, but indicate that viscosity 
is greatly reduced by dissolved gases such as water vapor; also 
that medium to basic magmas are less viscous than acid magmas. 
Rate of cooling.—Aside from its size, the temperature of the 
mass and of its walls affects the rate at which the magma passes 
1 Sydney Paige, ‘‘ Progressive Increase of Viscosity during Intrusion,” Jour. Geol., 
XXI, 541. 
