304 
of water may perhaps also raise the temperature of consolida- 
tion, and the chilling effect will thus be increased. 
The fact that batholitic magmas in several instances have 
failed to swallow up their cover is not, therefore, opposed to the 
hypothesis of intrusion by subsidence of roof blocks, and the 
value of the objection mentioned above, entirely depends on 
the correctness of the statement that batholitic magmas never 
break through. If, however, the possibility of overhead stoping 
be admitted, the ‘never breaking through’ cannot be considered 
as an established fact. Of course we cannot assume that an 
acid magma would appear at the earth surface as a lava lake 
like that of Kilauea. But when we do consider the probable 
result of a breaking through, it will be found that certain ob- 
served phenomena are not at all incompatible with the suppo- 
sition that many batholitic magmas may have actually penetrated 
their cover, and some may perhaps have done this on more 
than one occasion. 
Penetration of the cover would presumably lead to a vol- 
canic outburst of catastrophic character with more or less 
heavy explosions, accompanied by the outpouring of heavy lava 
flows and the formation of a solid lava crust. The immense 
loss of heat, caused by such an outburst, would establish a 
period of quiescence, but after a time hot magma from below, 
being brought by convective currents into contact with the 
newly formed roof stoping, would be set up again. Thus inter- 
rupted, perhaps by repeated volcanic outbursts, the stoping pro- 
cess would continue until the dying away of the igneous acti- 
vity. The final result would probably be a batholite under a 
relatively thin cover of its own lava, and at the contact the 
batholitic rock would present a fine-grained, marginal facies 
sending apophyses into the roof rock, without any direct evi- 
dence of the breaking through. 
But in many cases the record of the breaking through 
might be yet more concealed. If volcanic action continued at 
