^64 Dr Daubeny on the Volcanoes (^Auvergne, 
having been once the crater, from whence the lava of Voloic 
was ejected. It is interesting to remark, that the stream in its 
descent appears to have been arrested by a sort of knoll of gra- 
nite, which probably rose considerably above the general level, 
and by the .obstacle which it opposed to its progress, caused it 
to divide into two branches, between which this little granitic 
eminence still is seen protruding, — a solitary vestige of the stra- 
tum which formerly existed on the surface, but it is now over- 
spread with lava. The two branches of the main stream ap- 
pear to have become reunited below ; and having descended 
the slope of the hill, to have spread themselves over the valley 
of Volvic, extending oyer it to within a mile perhaps of the 
town of Riom. 
The only remarkable mineral found among the lava of ¥ 0 !- 
vie is specular iron-ore, or fer oUgiste of the French School. 
It occOrs in the crevices of the rock, as well as disseminated 
through its porous structure in minute plates of a bright metallic 
lustre. It is met with also in the Puy de la Vache, the Puy de 
Dome, and among the trachytes of Mount Dor ; and the same 
tnineral is a well known product of volcanoes now in activity, 
where it is generally considered to be the effect of sublimation. 
The lava of the Puy de Come, a mountain a few miles to the 
north-west of Clermont, which I unltickily omitted to examine, 
is described by M. Mont Losier'j the author of a little Essay oii 
the Theory of the Volcanoes of Auvergne, is equally interest- 
ing with that of Volvic. The lava that has flown from this 
hill divides, lie says, into two branches, one of which flows di- 
rectly into the bed of the River Sioule, whilst the other takes 
the direction of a place called Tournelisej reaches the village of 
Pont Gibaud, and terminates, like the other, by flowing into the 
bed of the same river, about three miles lower down. A tor- 
rent of this description might naturally be expected to effect 
singular changes in the face of the cotuitry which it traverses. 
Accordingly we shaU find that it has blocked up a little val- 
ley which formerly seems to have had a drainage to the west, 
on the side of Chambois and Massanges, and has converted it 
into a sort of swamp, known by the imposing name of Lac de 
Come. Lower down, the same stream of lava has occasioned 
still greater changes. The rivers Sioule and Menges former- 
