SHALER: EXPULSION OF GASES. 
97 
find its way slowly downward, so that it in time displaces the original 
water of construction to the extent above noted. There arises the 
question as to how the rain water can pass downwards without 
allowing the gas contained in the lower levels to find its way upwards 
by the same channels. The suggestion I would make is, that in part 
the descent of the pure rain water is due to osmose action, that 
curious process by which the contents of fluids separated by a permeable 
wall are equalized. It may be that in the osmose movement the upward 
transfer of the gaseous contents of the lower lying water is under 
conditions not effected in the same measure as is the diffusion of 
the fluid. There seems to be no recorded experiments which can 
be used to determine this matter; it appears to be worth further 
enquiry. 
Returning to the main question as to the ascent of gas-impelled 
fluids in vertical shafts, let us consider the possible bearing of the 
principles which may be observed in earthquake fountains and in 
springs upon the theory of volcanic action. It is evident that there 
is a general resemblance between the movements of gases from the 
superficial regions of the crust and those which come within volcanic 
explosions from the deeper realm. In both cases the impulse results 
in the formation of vertical shafts up which the materials are impelled; 
in both these lines of passage appear to be opened by the action of 
the gases; in both the times of escape are commonly occasional and 
the violent actions of this nature to a certain extent connected with 
the occurrence of sudden movements of the crust. Examining the 
evidence as to the character and modes of operation of the volcanic 
forces, we are brought to the hypotheses which we shall set forth below. 
It apj:>ears evident that all lavas at the moment when they reach 
the surface are intimately mixed with steam ; this vapor is present in 
large quantities ; though much of it is disengaged at the vent, enough 
often remains in the excurrent lava streams to make them cloud- 
wrapped for a distance of some miles from the place of exit. 
Although some geologists are of the opinion that the steam results 
from the contact of superficial or rain water which has penetrated to 
the roots of the volcano, the view seems untenable and this for the 
reasons which we may briefly note. In the first place, there appears 
to be no way by which the lava encountering bodies of water on its 
ascent could become commingled with that fluid, the result of which 
contact would it seems necessarily be to produce explosions which 
would send the steam forth uncommingded in an intimate way with 
