402 THE DEPTHS OF THE SEA. (CHAP. VI. 
depressions in the surface isotherms, the balance of 
probability seems to be in favour of the view that the 
conditions of temperature and the slow movement of 
this vast mass of moderately cold water, nearly two 
statute miles in depth, are to be referred to an 
Antarctic rather than to an Arctic origin. 
The North Atlantic Ocean seems to consist first of 
a great sheet of warm water, the general northerly 
reflux of the equatorial current. Of this the greater 
part passes through the Strait of Florida, and its 
north-easterly flow is aided and maintained by the 
anti-trades, the whole being generally called the 
Gulf-stream. This layer is of varying depths, ap- 
parently from the observations of Captain Chimmo 
and others, thinning to a hundred fathoms or so in 
the mid-Atlantic, but attaining a depth of 700 to 800 
fathoms off the west coasts of Ireland and Spain. 
Secondly of a ‘stratum of intermixture’ which ex- 
tends to about 200 fathoms in the Bay of Biscay, 
through which the temperature falls rather rapidly ; 
and thirdly, of an underlying mass of cold water, 
in the Bay of Biscay 1,500 fathoms deep, derived as 
an indraught falling in by gravitation from the 
deepest available source, whether Arctic or Antarctic. 
It seems at first sight a_ startling suggestion, 
that the cold water fillmg deep ocean valleys in the 
northern hemisphere may be partly derived from 
the southern; but this difficulty, I believe, arises 
from the idea that there is a kind of diaphragm at 
the equator between the northern and southern ocean 
basins, one of the many misconceptions which follow 
in the train of a notion of a convective circulation in 
the sea similar to that in the atmosphere. There is 
