212 
THE MEDITERRANEAN NAT! R A LIST 
have proportionally raised the boiling point. Then 
again, as the water exists bound up, as it were 
within the pores of some permeable rock, littie 
convexion circulation is permitted, at the same 
time that expansion to the gaseous condition 
furthermore is resisted. This shell of superheated 
water is only separated from the igneous magma 
by the salband, which according to varying cir- 
cumstances may differ very much in thickness, 
and so will act as a more or less imperfect porous 
septum between the igneous matter and super- 
heated water. Although probably not possessing 
exactly the same physical characters as the porous 
septum in dialysis, nevertheless it no doubt would 
permit diffusion to go on between the two fluids 
which it separates, or even the porous rocks them- 
selves may play that part. Besides, we have an- 
other striking resemblance to the process of 
dialysis, for the igneous magma is in a vitreous 
state, which we may take as the representative of 
the colloids, (1) whilst the superheated water in 
all probability may still be regarded as a crystal- 
loid. In consequence of this we should look for 
endosmosis as the principal function, although the 
metamorphism of surrounding rocks, which in the 
case of the existence of salbands is comparatively 
slight, would indicate some amount of exosmosis. 
In the case of the blind fissure being converted 
inti; a channel through which the igneous magma 
circulates, then no doubt the salband would, in 
most cases, be refused or carried away by other 
means, and the permeable rocks would then play 
the part of the septum. In fact, even in a blank 
fissure we can comprehend that no salband may 
exist. 
The rapidity of this endosmosis of water, and its 
diffusion through, or solution in, the colloidal-like 
magma, will obviously depend upon a variety of 
circumstances, such, for instance, as the composi- 
tion of the magma, the form of fissure, and 
therefore amount of surface exposed, pressure, «fec. 
This we see portrayed in the illustrations we chose; 
for if carbonic anhydride is in contact with the 
calni surface of water, solution takes place very 
slowly. A knowledge of this fact is practically 
made use of in the seltzer-water machine, in which 
a number of lashers revolve with great rapidity in 
(1) At any rate as Jar as the silica, and proba- 
bly the alumina and iron oxides , are concerned. 
a chamber filled with water and the carbonic 
anhydride, so that a very large surface of each is 
brought into contact by the churning motion, and 
consequently solution takes place with very great 
rapidity. 
But to return to the main question, this absorp- 
tion of water will go on at the expense of heat to 
the igneous magma, which, however, will only lose 
so much as will raise the amount of water absorbed 
to its own temperature. This loss will not. of 
course, be very great, since there is no conversion 
of a liquid into a gas. Nevertheless this loss of 
heat, combined with that due to the conduction 
away by the surrounding rocks, may be so great 
that the igneous magma may reach its point of 
solidification, and further action will be prevented 
by the fissure being now filled by a cooled rock 
mass. 
On the other hand, should such not have taken 
place, as the amount of water absorbed increases, 
the tension of the fluid mass will proportionally 
do so also. There will arrive a time when the 
tension of the fluid mass will exceed the resistance 
of the surrounding rocks, or the superincumbent 
pressure, which will result in the rending asunder 
of them and the extension of the fissure. Such 
extension may be sufficient to make it reach the 
surface forming the site of a volcano, or as it 
extends and gives place for expansion the tension 
may proportionally so decrease until the balance 
is restored before the surface is reached. The 
extension of such a fissure will rather tend towards 
the surface, as least resistance would be encoun- 
tered in that direction. We may thus have a dyke, 
a laccolite or a volcanic neck. Contact metamor- 
phism is very little in the first more in the second 
and well marked in the third if the volcano 
produced has been permanent so that the neck 
becomes a feeder canal. 
Such an extension of a fissure will give rise 
to two or more very distinct series of vibra- 
tions: first, we shall have slow ones extending 
over a considerable length of time, due to the 
gradually increasing compression around the ex- 
pansible matter which, if apparent at the surface, 
would assume the characters rather of tilt than 
that of an earthquake. Local elevation of a small 
area such as occurred at the Starza of i’ozzuoli, 
pending some years before the outburst and for- 
