10 PROFESSOR FORBES ON THE VOLCANIC GEOLOGY OF THE VIVARAIS. 
the lava wave, and thus, as it were, unrolls a continuous carpet, over which the 
more liquid stream flows; here and there breaking through the tenacious but 
partially yielding and crackling crust by which it is imperfectly confined.* This 
perfectly explains why we should have such a substratum of scorize in this place, 
which is often wanting below the basalt in the higher and steeper valleys where 
the lava flowed more rapidly, and it also accounts for the arched forms of the 
contact-surface, which fig. 3 remarkably displays; these vaults being due to the 
actual rolling of the lava over the more or less ponderous masses of scorize de- 
posited from its own surface. The brick-red colour common in scoriz beneath 
lava is to be attributed to the intense heat communicated to them by the lava, 
after they have been covered by it. This heat must have been retained for a 
prodigious space of time. Brick-red tints are usually produced upon minerals 
subjected to close or confined heat. The casual removal of the scorie has, in some 
places, left grottoes beneath the vaults. Secondly, These vaults also present this 
remarkable peculiarity, that the columnar structure of the lava (which here, 
as elsewhere, is best developed near its lower surface) conforms so nicely to the 
contact-surface with the scorize, to which it is always nearly perpendicular, as to 
give quite the appearance of vaulting stones to the covering of the grottoes. This 
is a fact so general amongst the lavas of the Vivarais, as to deserve almost to be 
called universal. It is interesting, as illustrating the development of the pris- 
matic structure which was so long supposed to distinguish ancient basalts from 
true lavas, but which is now universally admitted to characterize both, when the 
circumstances of cooling are favourable to their production. The fact of the per- 
pendicularity of the columns to the surface of cooling admits of this general 
illustration,—that if A (Plate II., fig. 4) represents a cold mass of rock over- 
flowed by lava, which gradually loses its heat by contact and conduction, all the 
points equidistant from the rock, as a, a, a, or b, b, 6, or ¢, ¢,¢, may be conceived 
to be, at the same moment, in the same condition as regards a tendency to con- 
solidate or to crystallize. Any peculiar action, which depends upon a particular 
stage of cooling, will therefore affect similarly all the points a, a, a, and so of the 
rest; that peculiar state of tension which produces the columnar division, will 
therefore prevail uniformly over any one of these isothermal surfaces (or surfaces 
equally cooled) at a given time, and will tend to produce its effect everywhere on 
that surface, and the lines or planes of separation will therefore seem to proceed 
uniformly from the surface of cooling in a direction perpendicular to it. Or on 
the less probable hypothesis of the columns being due to real crystallization, the 
crystals will naturally begin to form at the surface of earliest consolidation, 
* That such is the mode of progression of lava streams at a great distance from their origin, 
or after they have been running for a long time, appears from the descriptions of the best writers on 
voleanoes. Compare Scropr’s description of the lava of Etna of 1819, in his work on Volcanoes, 
p- 102; and Avtpso’s figure of the descending lava waves of 1831, in his Description of Vesuvius, 
p- 92. 
