If the frigate is intended to approach the small cluster of islands of Fernando de Noronha, E. of Pernambuco (Lat. 3° 50' S.), I would recommend to that excellent geognostic, Dr. Hochstetter, the hornblendic phonolithe rock found there, far from a volcanic crater, but with trachytic dykes and basaltic amygdaloid. The flat little island of St. Paul (Peñedo de San Pedro), 1° N. Lat., singular to say, is not volcanic at all, containing, like the Malouin or Falkland Islands, slaty green-stone passing into serpentine. Should the frigate alter her course and cross the Equator more to the eastward, without touching at Rio de Janeiro, she might possibly fall in with the Marine Volcanic region, (Lat. 0° 20' S., Long. 22° W.), which has quite lately become famous again by the U. S. Expedition of the Brig _Dolphin_ (1854), commanded by Lieutenant Lee. On 19th May, 1806, columns of black smoke were seen issuing from the sea by Krusenstern, and volcanic ashes were gathered, after a singular bubbling of the sea from 1748 to 1836, according to careful investigations by Daussy. As the frigate is commissioned to visit Ceylon and the Nicobar Islands, she cannot sail direct from the Cape to Australia; and the hope must therefore be abandoned of her visiting the small basaltic islands, known as Prince Edward's (47° 2' S., 38° E.), and Possession (46° 28' S., 47° 30' E.), belonging to the Crozet's Group, or the two islands, long confounded with each other, of Amsterdam (Lat. 37° 48' S.) and St. Paul (Lat. 38° 38' S.) The latter island, the more southerly of the two, (a very characteristic drawing of which was given by Willem de Vlaming so far back as 1696), is supposed to be volcanic, not only by its form, which will at once remind the geologist of Santorin, Barren Island, and Deception Island, (one of the New Shetland group), but also in consequence of the eruption of steam, and the flames occasionally observed there. As for Amsterdam, which consists of a single densely-wooded mountain, the puzzle remains for solution as to how, during the expedition of D'Entrecasteaux in 1792, the whole island seemed, during two entire days, enveloped in smoke; whereas, on landing there, the naturalists of that expedition were satisfied that the mountain was not an active volcano, and that the columns of steam issued out of the ground near the shore! As yet, the phenomenon remains entirely unexplained. If we examine any map of the Indian Ocean, we may trace the continuation of the Sunda group from Sumatra, N.W., through the Nicobar, and Great and Little Andaman Islands, and thence through the volcanoes of Barren Island, Narcondam and Cheduba, nearly parallel with the coasts of Malacca and Tenasserim, all on the eastern part of the Bay of Bengal. The minor volcanoes just enumerated will present valuable opportunities of geological enquiry.