354 
course), there are also others which both in age and structure 
occupy a position intermediate between the naujaite, properly 
so called, and the ordinary pegmatites. These intermediate 
types are strictly regulated in their distribution by the benching 
of the rock, and should perhaps be considered as the latest pro- 
ducts of the consolidation phase rather than as the first pro- 
duets of the pneumatolytic phase. It is probable that the 
sodalite-foyaite in the manner of its formation ought to be com- 
pared with these. 
Under this assumption the succession of events leading to 
the formation of naujaite and sodalite-foyaite is believed to have 
been as follows. 
When the crystallization of the agpaitic magma commenced 
fractional crystallization and gravitational separation caused an 
accumulation of sodalite crystals in the uppermost portion of 
the magma. If consolidation had been accomplished without 
reduction of volume the entire upper portion of the magma 
would have crystallized as naujaite, and the sodalite-foyaite 
would not have come into existence. But as crystallization in- 
volved contraction water and other expansible matter would 
tend to accumulate above the naujaite; and if we assume that 
volatile substances lower the temperature of consolidation and 
the specific gravity of the magma, it is evident that by degrees 
a moderate quantity of magma rich in volatile matter would 
accumulate above the naujaite (and later between the naujaite 
benches also). This uppermost magma sheet by reason of its 
low specific gravity would prevent sodalite crystals from rising, 
and perhaps by this time the temperature would have fallen so 
much that viscousity hindered further differentiation. Under these 
conditions the magma would not grow richer in sodalite, and 
when consolidation took place the structure would be similar to 
that of the pegmatite benches mentioned above. In this way 
both the position, the composition, and the structure of the 
sodalite-foyaite would be explained. 
