principal one. There are also several volcanic cones, apparently of this
modern period, round the circumference of the island, especially at the
northern end, where they form separate islets.

The mountains composed of the more compact and crystalline basalt, form the
main skeleton of the island. M. Bailly ("Voyage aux Terres Australes" tome
1 page 54.) states that they all "se developpent autour d'elle comme une
ceinture d'immenses remparts, toutes affectant une pente plus ou moins
enclinee vers le rivage de la mer; tandis, au contraire, que vers le centre
de l'ile elles presentent une coupe abrupte, et souvent taillee a pic.
Toutes ces montagnes sont formees de couches paralleles inclinees du centre
de l'ile vers la mer." These statements have been disputed, though not in
detail, by M. Quoy, in the voyage of Freycinet. As far as my limited means
of observation went, I found them perfectly correct. (M. Lesson, in his
account of this island, in the "Voyage of the 'Coquille'," seems to follow
M. Bailly's views.) The mountains on the N.W. side of the island, which I
examined, namely, La Pouce, Peter Botts, Corps de Garde, Les Mamelles, and
apparently another farther southward, have precisely the external shape and
stratification described by M. Bailly. They form about a quarter of his
girdle of ramparts. Although these mountains now stand quite detached,
being separated from each other by breaches, even several miles in width,
through which deluges of lava have flowed from the interior of the island;
nevertheless, seeing their close general similarity, one must feel
convinced that they originally formed parts of one continuous mass. Judging
from the beautiful map of the Mauritius, published by the Admiralty from a
French MS., there is a range of mountains (M. Bamboo) on the opposite side
of the island, which correspond in height, relative position, and external
form, with those just described. Whether the girdle was ever complete may
well be doubted; but from M. Bailly's statements, and my own observations,
it may be safely concluded that mountains with precipitous inland flanks,