634 
MR. J. PRESTWICH ON SUBMARINE TEMPERATURES. 
Antarctic seas to the equator, and which no doubt has justly been attributed to deep 
undercurrents carrying the waters of the poles to tropical regions, and the probable rise 
of these polar waters to the surface in the equatorial zone of the Atlantic. The 
source of those glacial waters in the North Atlantic lies, probably, in the Arctic Ocean ; 
and the question arises as to the channels by which they travel southward. The 
comparatively high temperature of 34° to 36° at depths in the seas around Spitzbergen 
shows that, although a deep body of cold water may move down the east coast of Green- 
land, the channels of the comparatively shallow sea between Norway and Spitzbergen 
are entirely, and of the deeper sea between Spitzbergen and Greenland in great part, 
occupied by a body of warmer water from the south (for without renewal the degree ot 
heat could not be maintained), pn the other hand, the constant low temperature at 
depths in Baffin Bay, and the southward drifting of the large low-sunk icebergs, show that 
that sea and Davis Strait afford a passage to a deep glacial current derived from the Arctic 
seas of North America. Issuing from these comparatively narrow channels this body 
of cold water unites with that passing down the east coast of Greenland, and flows 
southwards, over the great depths of the Atlantic, apparently to the equator. 
In the South Atlantic, on the contrary, the channel of the deep-seated glacial water 
is coextensive with the wide expanse open to the Antarctic seas, so that an unbroken 
undercurrent of such waters may occupy the one broad bed of that ocean. 
These two great undercurrents of the Atlantic, flowing respectively from the north 
and the south poles towards the equator, must eventually meet ; and, judging from the 
rise of the bathymetrical isotherms and the low temperature of the sea immediately 
beneath the heated surface-waters in the equatorial regions, it is probable, as suggested 
by Lenz, that the meeting is there, and that it is that which in part determines, in 
conjunction with the excessive evaporation, the surging-up of the polar waters, though 
other causes presently to be referred to may assist. In whatever way effected, the 
waters which thus rise to the surface in the equatorial zone necessarily tend to disperse 
and escape into other areas, whether by a slow movement in mass, or by more rapid 
currents in shallower and more definite channels, or by both causes combined. 
The course of these deep Arctic and Antarctic undercurrents or streams in the Atlantic 
may be influenced by another cause ; viz. by the west to east trend of the South- 
American continent from the Caribbean Sea to Cape St. Roque, and by that from east to 
west of the African continent along the coast of Guinea — projections which both 
contract the width of the Atlantic, and present barriers which may help to deflect side- 
ways and upwards, on the one (American) side the southward flow of the Arctic waters, 
and on the other (African) side the northward flow of the Antarctic waters, in a 
manner analogous to that which takes place on shoals and islands. 
It is not my intention to enter upon the discussion of the course and magnitude of 
the Gulf-stream ; but I would suggest whether or not the initial start of that great 
current, together with the others which originate or acquire new power at the equator, 
such as the Guinea, the South Equatorial, and the Brazilian currents, may not be cradled 
