488 FRANK F. GROUT 
that a magma at 1,300°C. may be intruded at a horizon a mile 
below the surface which may have a temperature not over 100, 
average probably 50°. The heat loss can be calculated by the 
formula 
RAT —Ty) 
C= Gate dame? 
where & is the conductivity in C.G.S. units, A the area in square 
centimeters, and (7 —7;) the difference in temperature. So long 
as the surface of the earth was kept cool, the upward flow of heat 
would be fairly uniform until the magma cooled appreciably. On 
the other hand, the temperature at the floor will not remain uniform. 
The floor of the Duluth gabbro must have sunk nearly to miles. 
The isogeotherms would rapidly rise. The lava streams that fed 
the intrusion and even the approach of the magma chamber below 
would contribute to the rapidity of the rise. With these supplies 
of heat from below, the heat added to the floor by the gabbro would 
accumulate rather than pass on. A rough calculation by the same 
formula indicates that the gabbro added heat to its floor in a year 
equivalent to a general rise in temperature of 12° for the first mile. 
In 100 years the floor would be so hot that heat losses in that 
direction would be very small. The gabbro as a whole, however, 
would cool only a few degrees in too years. Thus it seems that by 
the time crystallization begins the loss of heat is likely to be from 
the top and sides of the chamber. If this loss occurred without 
convection—say from a zone to to 15 meters thick at the roof— 
this zone would be cooled toc” below the main body of the magma 
very soon, say within a month. In a few years the contrast in 
temperature would be much greater. It is therefore assumed that 
a difference in temperature of roo°C. in different parts of a medium 
to large magma is not unusual. 
The specific gravities of phases ——Specific gravity varies with 
composition, temperature, and crystalline or glassy structure. 
Daly, in summarizing the results of several investigators, estimates 
that the change from liquid to crystalline rocks at the same tempera- 
ture results in an increase of specific gravity as follows: 6 per cent 
for gabbros and diorites, 7 per cent for quartz-diorite, 8 per cent 
