102 The American Geologist. February, 1904. 
should ever cease. But the new hypothesis is more favorable 
to limited liquidity, since it favors heterogeneity of the earth's 
mass and consequent local variation in temperature and melt- 
ing point. 
The varied phenomena commonly known under the term 
"volcanism" are much better explained by the new hypothesis. 
It is evident that the explosive action in volcanoes is due to the 
expansive power of heated vapors, chiefly water. Under the 
old view the supply of water is from the hydrosphere, but the 
assumption that volcanic water is meteoric is founded mainly 
in analogy and has little basis in observation or sound reason. 
We may admit that the water of thermal springs and geysers 
and perhaps some fumaroles is atmospheric. But the volcanic 
water is too large in quantity and apparently from too great 
depths to be derived from superficial sources. The amount of 
volcanic water is enormous. Fouque determined that the 
amount of steam expelled from one of the numerous parasitic 
cones of Etna was equaled in 100 days to 462 million gallons of 
water. This equals 16 gallons for every square foot of a 
square mile, or a depth of 32 inches over that area. But the 
steam product of the single cone was probably not the one 
thousandth part, perhaps not the one ten-thousandth part, of 
the whole product of the volcano during that time. 
The interstitial water of the rocks can not be sufficient to 
product the vast accumulations localized in the volcanic reser- 
voirs, and efficient circulation at great depths seems impossible. 
The capacity of molten magmas to absorb vapors has been used 
as an argument for the meteoric source of the water, but an ob- 
jection is that before the water can reach the absorbing magma 
it must pass through a great thickness of moderately heated 
strata where the repulsive force is great and the absorbing 
power is small or entirely wanting. The idea of supply 
through fissures, subterranean or oceanic conduits, in the face 
of the vastly superior and opposing hydrostatic pressure of the 
molten rock is not worthy of consideration. 
The presence of water and other gases in the crystalline and 
deep-seated rocks is a familiar fact to petrographers, and the 
theory that the volcanic waters were indigenous to the earth's 
magma has, in later years, been held by several students of 
volcanism, for example, Scrope, Fisher, Reyer, Tschermak 
