deposited, is obscure; but the fact has been observed in several other
places. (Spallanzani, Dolomieu, and Hoffman have described similar cases in
the Italian volcanic islands. Dolomieu says the iron at the Panza Islands
is redeposited in the form of veins (page 86 "Memoire sur les Isles
Ponces"). These authors likewise believe that the steam deposits silica: it
is now experimentally known that vapour of a high temperature is able to
dissolve silica.) In some half-decayed specimens, I found small, globular
aggregations of yellow hyalite, resembling gum-arabic, which no doubt had
been deposited by the steam.

As there is no escape for the rain-water, which trickles down the sides of
the ravine-like hollow, whence the steam issues, it must all percolate
downwards through the fissures at its bottom. Some of the inhabitants
informed me that it was on record that flames (some luminous appearance?)
had originally proceeded from these cracks, and that the flames had been
succeeded by the steam; but I was not able to ascertain how long this was
ago, or anything certain on the subject. When viewing the spot, I imagined
that the injection of a large mass of rock. like the cone of phonolite at
Fernando Noronha, in a semi-fluid state, by arching the surface might have
caused a wedge-shaped hollow with cracks at the bottom, and that the rain-
water percolating to the neighbourhood of the heated mass, would during
many succeeding years be driven back in the form of steam.

TAHITI (OTAHEITE).

I visited only a part of the north-western side of this island, and this
part is entirely composed of volcanic rocks. Near the coast there are
several varieties of basalt, some abounding with large crystals of augite
and tarnished olivine, others compact and earthy,--some slightly vesicular,
and others occasionally amygdaloidal. These rocks are generally much