Natural History of Volcanos and Earthquakes. 257 
tures, by reason of their unequal specific gravities, seems to be 
very much interrupted by the narrowness and sinuosity of the 
passage. The intermitting springs of Iceland are probably 
caused by the existence of caverns, in which the vapor is re= 
tained by the pressure of the column of water in the channel 
which leads to the surface. Here this vapor collects, and presses 
the water in the cavern downwards until its elastic force becomes 
sufficiently great to effect a passage through the column of water 
which confines it. The violent escape of the vapor causes the 
thunder-like subterranean sound, and the trembling of the earth, 
which precede each eruption. The vapors do not appear at the 
Surface till they have heated the water to their own tempera- 
ture. When so much vapor has escaped that the expansive 
force of that which remains has become less than the pressure 
of the confining column of water, tranquillity is restored, and 
this lasts until. such a quantity of vapor is again collected as to 
produce a fresh eruption. The spouting of the spring is, there~ 
fore, repeated at intervals, depending upon the capacity of the 
cavern, the height of the column of water, and the heat gene- 
tated below.* 
The two distinct classes of eruption in the Geyser, which we 
have already mentioned, seem to be attributable to two different 
cavities. A smaller cavity fills quicker, and, therefore, empties 
itself more frequently ; a larger one fills slower, empties itself 
seldomer, but with greater violence. But the playing of the 
Geyser, the Strokr, and some others, is subject to very great va- 
i 
r in this state of activity, It 
then again became quiescent, ‘except that the water, deep in the tube, continued, 
as Usval, to boil violently. 
Vol. Xxxv1, No. 2.—April-July, 1839, 33 
