114 Mr. Dana on the Analogies between the 
kernels there is a narrow band of nearly pure epidote about an 
inch and a half thick. 
Many other similar examples are minutely described by Prof. 
Rogers in the same highly interesting report. In one, where the 
sandstone was altered to the same distance—about one fourth of 
a mile—the altered rock within fifty yards of the dyke, contained 
thickly disseminated crystals of tourmaline, some of them half 
an inch in diameter. 
It is surely impossible that by conduction alone the heat of the 
ejected dyke should have been conveyed to so great a distance 
from the dyke, and in such a degree as to produce crystals of 
tourmaline and epidote, the latter one fourth of a mile from the 
source of heat. These effects must be imputed to the heated 
waters rendered hot by the eruption. The compact structure of 
the trap leaves no doubt that the eruption was submarine. 
A moment’s consideration of the circumstances attending such 
an eruption will place the subject ina clearer light. ‘The sub- 
marine sedimentary deposit, whether a bed of sand or clay, is 
soaked with water; and between its layers, or in the cavities or 
caverns interspersed through or between these submarine beds, 
the waters are collected in large quantities. As the fissure opens, 
the melted rock flows up from below to fill it; the interspersed 
or permeating waters are heated by its sides and convey the heat 
far into the rock. ‘The ocean’s waters too enter the fissure as 
soon as opened, and meet the liquid fires on their ascent; hot, and 
in commotion from the ignition and violent ejection of the fluid 
rock, the waters are thrown into any open cavities in the walls, 
and thus aid in diffusing the volcanic heat. 'The superincumbent 
waters are next heated, and currents intensely hot spread around 
by the attendant convulsions, diffuse the heat far and wide over 
the surrounding sedimentary deposits, which are thus permeated 
and buried in the fluid heat. ‘Through such influences, we may 
account for all the alterations and crystallizations above described. 
The effects of the hot waters may probably be seen beyond 
the hardened portion of the rock, in the red color of the sand- 
stone. We know that this is a common effect in modern volca- 
nic regions, and can detect the same in many more ancient. 
The blue and purple colors which the altered rocks assume 
near the dyke, arise, as in the common burning of bricks, from 
the excessive heat, which in part deoxydizes the iron or enables 
