332 
lying rocks subsided during its crystallization it must have 
been otherwise with the lujavrite. The cause of this peculiar 
condition may perhaps be found in the very low consolidation 
temperature of the lujavrite. We do not know what underlies 
the lujavrite at Ilimausak, but whether it be an essexite, or an 
augite-syenite, or any other rock, it had probably a higher 
consolidation temperature than lujavrite, and it may thus in 
spite of the greater depth have been capable of crystalliz- 
ing first. 
By these assumptions — (1) a low temperature of conso- 
lidation; (2) a large proportion of water etc.; and (3) a solid 
substratum — the peculiar structure of the lujavrite seems to 
be intelligible. The subsidence and the fissuring of the upper 
strata, and the accompanying forcing up of a portion of the‘ 
lujavritic magma along the eastern and perhaps the northern 
borders of the batholite, would facilitate the escape of volatile 
substances, and would thus accelerate the cooling and crystal- 
lization, but would counteract the development of large crystals. 
The solid substratum would prevent the supply of additional 
magma, and the contraction produced by the escape of water 
vapours, and by the transition to the crystalline state must 
therefore have resulted in a subsidence of the overlying rocks 
which continued during the entire period of crystallization. 
This subsidence of the solid roof, being somewhat irregular, 
would produce movements within the magma basin, and the 
movements would continue, and gradually affect all parts of 
the magma, until even the deepest portions had contracted to 
solid rock. The movement, thus, would be, not only an addi- 
tional cause favouring the development of a fine-grained struc- 
ture, but might also produce a primary schistosity of the 
whole mass. The observed geometry of the schistosity, the 
dip of the parting planes etc. are in agreement with this hypo- 
thesis. 
It will easily be seen that the above attempt to explain the 
