98 PROCEEDINGS: BOSTON SOCIETY NATURAL HISTORY. 
the molten rock. Such conditions, as we may observe when many 
cubic miles of lava are suddenly cast forth in a state where the mass 
is intimately commingled with the vapor, must be accounted for before 
this hypothesis can be accepted. In the second place, the general 
distribution of volcanoes active and extinct shows very clearly that 
their formation depends upon the continuance of some actions which 
go on beneath the level of the sea. So far as the downward penetra¬ 
tion of water goes, this is as effective beneath the land areas as it can 
be below the level of the oceans. Therefore, if the explosive 
phenomenon of volcanoes depends upon the contact of lava with water 
near the point of exit, there is no reason why the retreat of 
the sea from a field of their activity should be followed by their 
extinction. While it does not seem possible to explain the origin or 
the nature of volcanic explosions or of the materials which they eject 
by the supposition that the only contact of lava with water occurs in 
the upper part of the earth’s crust, the difficulty in accounting for the 
phenomena disappears as soon as we suppose that the process of 
ejection depends upon the vaporization of the construction water of 
the rocks whence the lavas are derived. As the subsequent considera¬ 
tions of this paper depend in large part upon the view which is to be 
taken as to the origin and nature of volcanic explosions, a summary 
of a part of the facts and of the conclusions which may be derived 
therefrom needs to be given. 
A study of the distribution of active volcanoes shows that they are 
essentially marine phenomena; they occur in a widespread manner 
and of a sufficient height occasionally to overtop the surface of 
the sea in all the oceans and in many of its lesser embayments; 
this essentially marine field of the vents at some parts over¬ 
laps the land but probably at no point for more than about two 
hundred and fifty miles from the shore. The vents classed as 
active which are more than one hundred miles from the shore 
line are languid, their outbreaks are of slight energy and very 
infrequent. If the sea deserts a volcanic district, the extinction 
of the explosive forces appears in all cases to be brought about. The 
facts on which these statements are founded are so numerous and of 
such wide geographical distribution that they may be accepted as a 
guide beyond the chapter of accidents. Taken with the connection 
of the intimate mixture of steam with lava they lead to and are a 
verification of the following hypothesis: — on the ocean floor sedi¬ 
ments are constantly accumulating ; although the rate of this accumu- 
