SHALER: EXPULSION OF GASES. 
103 
Shortly after I left the edge of the crater, as I was compelled to 
do by the falling away of the wind, which exposed me to the vol¬ 
canic bombs, I noted that the exjdosions increased in frequency 
until the discharge appeared, as far as the eye and ear could discern, 
to be continuous, yet as the succession of these reports was evident 
until their intervals became so small as to escape attention, I con¬ 
clude that even when the discharge appears to be continuous it is 
actually made up of repeated explosions. The phenomena reminded 
one of what we may note in the jets of steam from a locomotive 
when it moves with increasing speed; at first we notice the succes¬ 
sion of escapes but gradually as the intervals become shorter the 
action appears continuous. 
If the escape of steam from a volcano is in the bubble form, we 
have yet another likeness of the action to the movement of gas in 
the case of earthquake fountains. We may now suggest that in 
the volcanoes, as in the other instances of gas escape, the formation 
of these vesicles determines the construction of the vertical shaft 
and that the procession serves to maintain the opening of that shaft. 
On the basis above suggested we may regard the formation of a 
volcano as beginning with an aggregation of vesicles of steam which 
has taken place along some line of weakness. The weakness may 
be relatively slight, for even in as homogeneous a material as water 
there are differences of resistance sufficient to determine the paths 
of the bubbles. As these bubbles accumulate they diminish the 
resistance of the overlying mass, they push it aside, perhaps in part 
melt it by the heat which is conveyed upward along the paths of 
movement, until the time comes when the superincumbent material 
is blown away. 
Hitherto we have endeavored to account for the occurrence of 
volcanoes at particular points on the supposition that their channels 
were determined by the existence of faults. While such lines of 
weakness probably serve as directories in guiding the ascent of the 
vapors, there are many volcanic pipes which cannot reasonably be 
explained on the supposition that they have followed pre-existing- 
fault lines. There seems to be no more reason for requiring the 
intervention of distinct ruptured crevices in the case of volcanoes 
than there is in the case of the lesser but perfectly comparable 
shock fountains. We have only to suppose the development of 
pressure sufficient to break away the resisting covering and then to 
bring in the principles of the processions of bubbles to account for 
the observed facts. 
