226 THE ATLANTIC. [ CHAP. III. 
differences, almost within the limit of error of observation, may 
appear somewhat exaggerated. 
One marked feature in this section is the comparative thin- 
ness of the surface layer of heated water, and the enormous 
thickness of the mass of underlying cold water, varying little in 
temperature throughout its depth; and a second is the great 
regularity of the curve of fall of temperature from the surface 
to the bottom (Fig. 55). The bottom temperature of the west- 
ern basin of the Atlantic is slightly but very decidedly lower 
than that of the eastern at corresponding depths. 
One point connected with some of our earlier observations is 
of interest in its bearing upon the interpretation of the temper- 
ature observations of the Porcupine cruises of 1869 and 1870. 
Curve A, Fig. 56, represents the vertical distribution of heat in 
the Bay of Biscay determined by the Poreupime in 1869, and 
Curve B the distribution of heat derived chiefly from bottom 
temperatures off the coast of Portugal. The marked peculiar- 
ity in these curves is a rise or “hump” between 200 and 900 
fathoms, indicating a temperature in that particular stratum of 
water from some cause abnormally high. 
This has been accounted for by the “ banking-down” against 
the coasts of Europe of the Gulf-stream, the north-eastern re- 
flux of the equatorial current; and it is naturally most marked 
at that point where the impact, as it were, of the Gulf-stream 
is most direct—the coast of the Lusitanian peninsula. 
It is indicated on all current charts that about the 40th par- 
allel of north latitude a great part of the Gulf-stream bends 
southward along the coast of Africa, a portion of it curving 
round and rejoining the equatorial drift, and another portion 
joining the Guinea current. Some of the observations in the 
present section appear to confirm in a remarkable way the view 
that this irregularity in the curve is due to the cause to which 
it has been attributed. The southern branch of the Gulf- 
stream is not impeded by land, and there is little or no “ bank- 
