438 
Proceedings of the Royal Society 
bottom. To seaward, in 1800 or 2000 fathoms, the bottom is a fine 
Globigerina ooze, which gets coarser and sandier as the water shoals 
in the channel, nntil, on the summit of the ridge, there is generally 
no deposit at all, and the bottom is rock or coral often coated with 
black oxide of manganese. 
An essential condition of the existence of coral, whether in the 
deep sea or in the shallow water of the atolls of tropical seas, 
seems to be circulation of the water. This serves the double pur- 
pose of bringing food and removing sediment. In shallow water 
this circulation is produced by the breaking of ordinary storm waves 
on the shore or on shoals; in deep water it is produced by the trans- 
formation of the tidal wave into a current on meeting with obstruc- 
tions such as ridges or banks, even though they may have a thou- 
sand or more fathoms of water on them. In both cases the current 
is produced in the same way, namely, by the annihilation of a 
wave. 
In Mr Murray’s theory of the formation of coral islands, the 
growth (in height) of a shoal, owing to the greater amount of sedi- 
ment which reaches it from the surface compared with the amount 
which can reach an equal area of ground at a greater depth, is an 
important feature. We have seen, however, that it is probable 
that, wherever the bottom rises in the path of the tidal wave, a 
portion of the wave is transformed into current. This current 
would naturally keep the summit of the shoal clean swept of all 
freshly falling sediment, and would thus tend to limit the growth 
of such a shoal. 
But this very agency must favour the existence of such animals 
as deep-sea corals on the areas kept clear of sediment. It seems 
reasonable, therefore, to expect that such areas would be occupied by 
deep-sea corals. These would not be likely to extend laterally 
beyond the cleared summit of the rising, but would grow upwards, 
the living resting on the debris of the dead. The tendency would 
be to raise a massive pillar with perpendicular sides up through the 
water, and it would grow until the conditions, principally of tem- 
perature, set a limit. In the “ Coral patch ” we have undoubtedly 
such a structure, the base of which rests on the crest of a hill 800 
fathoms below the surface. Its summit is now rather over 400 
fathoms below the surface, and it is in full growth. Judging from 
