62 4% ; REPORT—1858. 
(nearly one-half the area of Africa) bounded on the north by Japan, and on 
the west by the Philippines. 
From lat. 30° S., a sub-oceanic crest-line of shallows appears to spur off 
eastward from the volcanic foci of New Caledonia and New Zealand, and, 
trending westward and a little northward through the Tonga, Society, Mar- 
quesas, and Gallapagos Islands, connected by continuous banks, joins the Cen- 
tral American group of volcanoes, thus cutting the great ocean basin 
nearly into two secondaries, each of which is probably in a less marked man- 
ner subdivided,—the northern sub-basin, by a line through Christmas and 
the Sandwich Islands, to some point of the volcanic group of the Audrea- 
nofsky Islands in the Atlantic Archipelago, making in its course a wide 
sweep to the east and north through an almost continuous chain of isles and 
banks; and the southern sub-basin by a line from the Society Islands 
through Easter Isle and Juan Fernandez, and combining with the great 
Chilian volcanic chain at its eastern extreme. 
A vast fissure (noticed by Humboldt), and marked by an almost continuous 
line of volcanic vents, extends in a direction nearly east and west, right across 
Mexico, between lat. N. 18° and 19°. It is nearly 500 miles in length. Its 
main direction, if produced, bears upon the voleanic island of Revillegigedo, 
and, as Humboldt also thinks, probably extends to Mouna Roa, in the Sand- 
wich Islands. The Mexican extremity of this enormous crevasse probably 
marks the continental end of one of the great dividing ridges of the sub- 
basins of the Pacific. 
Within the great Pacific Basin will be found (tinted blue) most of those 
great areas of probable subsidence indicated by Darwin*. These bands will 
be observed occupying the great sub-basins of the ocean, not very distant 
from great volcanic lines, and although not (with our present imperfect 
knowledge of soundings) quite free from the suspicion of occasionally inter- 
secting such lines (e. g. Marquesas and Society Islands, Ladrone, and New 
Guinea), yet, on the whole, keeping surface positions intermediate to the 
volcanic cinctures adjoining or around them. 
Less distinctly we may trace the cincture of mountain- and volcanic chain 
around the shallower Atlantic basin, and, through it, upon the submarine 
elevations dividing its sub-basins. Thus, starting from Iceland; the Ferro 
Isles, Scotland, and the mountains of Wales and England (with the breach 
of the English Channel, a narrow line in relation to the scale of our present 
survey), the Rhenish-German chains, the French and Western Alps, the 
Pyrenees, to Cape Finisterre and the coast of Portugal, connect by the 
Azores, and by innumerable submarine rocks and shoals, across to New- 
foundland. Here the lines to the northward may be pronounced unknown, 
until, returning back to Iceland, we find it approximates to the point we 
left through the great igneous and abrupt coast-line of Greenland. 
In connexion with this oceanic basin, we have two probably subsiding 
Haste of land—the one in Davis's Straits, the other in the Baltic—both tinted 
ue. 
The Central Atlantic forms a well-marked basin girded with voleanoes and 
mountain-ranges. Leaving the last stated boundary-line at Newfoundland, 
and going again eastward to the Azores, thence through Madeira to the 
Canary Isles, the Cape de Verds and including the great sub-oceanic vol- 
canic region between 15° and 30° long. W., and lat. 3° N. to 10° S., going 
westward by the island of Fernando Noronha to Cape St. Roque on the ex- 
treme east of the South American continent, returning to Newfoundland, 
* See Dana on Areas of Subsidence in the Pacific. Ass. Amer. Geol., Albany, 1843, and 
Edin. Phil, Journ. (New), vol. 35. p. 341. . DS's yO Se 
<1 — 
