212 THE ATLANTIC. (CHAP. In. 
rapidly decreased in number on the surface and at the bottom 
as we progressed southward. The proportion of their remains 
in the globigerina ooze near the Crozets and Prince Edward Isl- 
and was comparatively small; and to this circumstance the ex- 
treme cleanness and the unusual appearance of being composed 
of Globigerine alone were probably mainly due. We found the 
same kind of ooze, nearly free from coccoliths and rhabdoliths, 
in what may be considered about a corresponding latitude in 
the north, to the west of Farée. 
The next seven soundings, extending along the section to 
a distance of about 1500 miles from Teneriffe, and at depths 
varying from 3150 to 2575 fathoms, are marked on the chart 
“red clay.” According to our present experience, the deposit 
of globigerina ooze is limited in the open oceans—such as the 
Atlantic, the Southern Sea, and the Pacific—to water of a cer- 
tain depth, the extreme limit of the pure characteristic forma- 
tion being placed at a depth of somewhere about 2250 fathoms. 
Crossing from these shallower regions occupied by the ooze 
into deeper soundings, we find universally that the calcareous 
formation gradually passes into, and is finally replaced by, an 
extremely fine pure clay, which occupies, speaking generally, 
all depths below 2500 fathoms, and consists almost entirely of 
a silicate of the red oxide of iron and alumina. The clay is 
often mixed with other inorganic matter, particularly with par- 
ticles, graduating up to the size of large nodules, of peroxide 
of manganese; and in volcanic regions, or in their neighbor- 
hood, with fragments of pumice. The transition is very slow, 
and extends over several hundred fathoms of increasing depth ; 
the shells gradually lose their sharpness of outline, assume a 
kind of “rotten” look and a brownish color, and become more 
and more mixed with a fine amorphous red-brown powder, 
which increases steadily in proportion until the lime has al- 
most entirely disappeared. This brown matter is in the finest 
possible state of subdivision, so fine that when, after sifting it 
