MICROPEGMATITE IN PURCELL SILLS. 
27 
(3) No shatter contacts were seen in any of the sills. 
(4) The heat consumed in such a series of reactions, 
which are necessary under this hypothesis, would be very 
great and it is doubtful if the magma contained such a 
quantity of heat. 
The remaining hypothesis is, that the granite (micropegmatite) 
in the composite sills was formed by differentiation in place. 
This hypothesis is the one favoured by the author. In this 
case, there was an intrusion of a magma whose composition, 
in the case of any single sill, was the average of the whole con- 
tents of the sill in question. These sills have in some cases 
basic upper and lower contacts. The homogeneous magma 
which filled the composite sills would cool first along the contacts 
and hence the upper and lower contacts of the sill would be basic. 
Differentiation, acting on the remaining fluid portion, would 
cause the heavier constituents under the influence of gravity to 
gather in the lower part of the chamber concomitantly with the 
collection of the acid differentiate towards the upper part of the 
chamber. A similar distribution of acid and basic material in 
a laccolith has been described by Pirsson. A vertical section 
across the Shonkinsag laccolith 1 is as follows : — 
Rock type. 
Thickness 
in feet, 
(centre) 
Thickness 
in feet, 
(outer wall) 
Leucite basalt porphyry 
5 
10-15 
Dense shonkinite 
5 
Shonkinite 
5-6 
Transition rock 
3 
Syenite 
25-30 
Transition rock 
Shonkinite 
15 
60-70 
75 
Leucite basalt porphyry 
15 
15 
140 
100 
The combined effect of convection and crystallization is used 
to explain the above occurrence of rock types in the laccolith. 
The same result could be obtained by differentiation under the 
ipirsson, L, V. ( U. S. G. S. Bull 37. 
