Natural History of Volcanos and Earthquakes. 245 
also a forerunner of almost all eruptions, especially of those of 
Vesuvius. 
But if a reaction should take place in the column of water, yet 
the rising vapor would soon be so far cooled down as to become 
liquid again, without the expansive force of the enormous quanti- 
ties of vapor formed at the volcanic focus being thereby percepti- 
bly diminished. In addition to this, the hydraulic resistance in 
the narrow channels, through which the water is admitted, in- 
creases very considerably as its velocity becomes greater. But 
the column of water, by which the aqueous vapor is cut off from 
communication with the surface, acquires very great velocity in 
those narrow channels, from the enormously increased elastic 
force of the steam, by which the resistance may very easily be 
increased to the extent of much more than 1000 atmospheres. 
So that, notwithstanding that the expansive force of steam whose 
temperature exceeds 1754° or 1881° F., is greater than the hy- 
ostatic pressure of the column of confining water, yet this re- 
Sistance may suffice, in the manner just mentioned, to raise a col- 
umn of lava, of even a greater height than we have above reck- 
oned, to the summit of the voleano. If we may be allowed to 
take a comparison with an analogous phenomenon, it may be re- 
Membered that the touchhole of a cannon, or of a bore-hole in a 
mine, does not weaken much the action of the powder, although 
Proportion of the diameter of the touchhole to that of the 
mouth of the cannon is as 1 : 30.* If Perkins’s well known ob- 
“ervation,t that water and steam cannot be forced through nar- 
TOW openings in the red-hot generator of a steam engine, is appli- 
Cable to the gigantic generator, which formed the volcanic focus: 
this might be added to the causes already mentioned, which afford, 
Tesistance in the channels through which the waters are admitted. 
8o long as the communication with the sea remained open, the 
Volcano could never come to a state of rest, although the forma- 
_ =a Ce ee Se 
Same, vol. xxv, p. 591. 
