ri 
— 
ON THE FACTS AND THEORY OF EARTHQUAKE PHENOMENA. 63 
we trace the line southwards through the several chains of the United States 
down into Georgia, where, with the comparatively narrow breach of Lower 
Florida, it is carried on by Cuba and the whole chain of volcanic islands of 
the West Indies to Trinidad and the South American continent again. 
The Gulf of Mexico and Caribbean Sea form a smaller but separate basin. 
In the southern Atlantic we can trace a dividing ridge through South 
Ascension—the great suboceanic tract just referred to—North Ascension, 
St. Helena, and probably to Cape Negro on the African west coast, and 
thence to the Cape of Good Hope, and returning westward by Tristan 
d’Acunha, thence S.W. to the Isle of Georgia (lat. 55° S.) and through the 
Falkland Islands to the volcanoes of the southern point of South America; 
but this, like the sub-basins, through the scattered indications which alone 
we yet have in the vast southern portion of the Eastern or Indian Ocean west 
of Australia, is uncertain. 
There is little doubt that Australia, on its northern existing coast-line, was 
once united with New Guinea and the Aru Islands west and south of it 
(Wallace, Silliman’s Journal, vol. xxv.), and possibly with much of the land 
outlying to the west of that vast and now isolated continent; if not, the 
intermediate seas would be much deeper than they are, and the west coast of 
Australia with its mountainous chains would bound an ocean basin whose 
western boundary would be marked by a line of volcanoes from New Guinea 
to New Zealand and the Southern Sea. 
The seas of Ochotsk, of Kamtschatka, of Japan, and, above all, the Chinese 
and Malayan Seas with Borneo in the midst, form so many distinct basins, 
small relatively to the vast areas we have been reviewing, but distinct and 
strongly marked. In the Chinese Sea we have a probable tract of subsiding 
land, tinted blue upon the- evidence of Darwin. The bay of Bengal, well- 
marked all round northward from Sunda, and belted with volcanoes tv the 
Ganges, and with mountains near the coast thence to Ceylon, joins probably 
Western Australia by a suboceanic ridge, indicated through the rocks of 
Greville and Compton, the Island of Apaluria with the adjacent submarine 
voleano of 1789, and the ocean shallows and soundings, about 100° W. 
long. and 20° to 25° S. lat. 
_ The separate basin of the Arabian Sea is equally distinct, from Cape 
Comorin along the Malabar coast, all highly mountainous, Beloochistan to 
the mouth of the Persian Gulf (itself a small basin), thence by the Arabian 
coast-line to the volcanic region at the mouth of the Red Sea, and into 
Abyssinia with its characteristic and enormous crater-form lake of Tzana 
(though as yet not possessing any earthquake record), and thence through 
regions scarcely known upon the East African coast, crossing to the Comoro 
Islands (volcanic) and to the mountainous regions of Madagascar,—the vol- 
canic islands of Bourbon, Mauritius and Rodriguez, the Nazareth and Saya 
banks, the Chagos Archipelago and the Maldive and Laccadive Islands, 
completing the cincture with the Malabar coast again. 
Along the great band of these islands, and thence trending westwards by 
the Saya bank, lies one of the great tracts of ocean-floor which Darwin has 
shown to be probably subsiding (tinted blue). Assuming that this really is 
a band of subsidence, it would be more probable that the volcanic girdle 
v 
takes a wider sweep to the south and west of this band, and, leaving the 
Island of Rodriguez, makes for the volcanic centre marked in the ocean at 
long. 90° E., lat. 10° S., and thence turns northward to join Ceylon, Cape 
Comorin and the volcanic region of Pondicherry. 
_ Leaving the great ocean and great continent, we trace smaller basins 
(or rather saucers, for their extreme shallowness in relation to their surface- 
area must never be lost sight of), where larger portions of the elevated moun« 
