154 ON THE PHYSICAL THEOET OF UNDER-CUREENTS. 



indicating, probably, Antarctic water. The next three sections 

 were devoted to investigating the Gulf stream proper, or Florida 

 current, and were taken from St. Thomas to Bermuda, Bermuda 

 to Halifax, and Bermuda to New York. They showed very 

 plainly how minute a sectional area was occupied by the Gulf 

 stream compared with the enormous body of water alleged to be 

 put in motion by it ; and the continuity of the " cold wall " on 

 the American coas^ with the cold polar underflow was pointed 

 out, and its rationale explained. The fifth section was from 

 Bermuda, through the Azores, to Madeira, and the sixth from 

 Madeira, through the Cape de Yerde Islands to a position in 

 lat. 3° N. and long. 15^ W., in which — although the surface 

 temperature was much warmer — the isotherm of 40*^ rose from 

 900 fms. depth at Madeira to 400 fathoms in the equatorial position. 

 The seventh section, which comprised the Equatorial Atlantic, 

 was taken from the last-named position through St. Paul's Rocks 

 and Fernando !N'oronha to Pernambuco, in lat. 7^° S. In this 

 section, the uprising of polar water to the 'surface was clearly 

 demonstrated ( 1 ) by the very rapid fall in the thermometer, from 

 78^ at the surface to 40^ at 300 fathoms depth, while in both N. 

 and S. Atlantic oceans, outside the tropics, that temperature was 

 not reached till about 800 or 1,000 fathoms, depth, and (2) by the 

 specific gravity observations. In the I^orth Atlantic the surface 

 water was about 1*0L72, and the bottom (or polar) water was 

 about 1-0263, while in the equatorial Atlantic the surface water 

 was 1-0263, and the bottom water nearly the same, 1 0261. In 

 this section also the bottom temperature fell to 32^^.4, showing the 

 derivation of this water from an Antarctic source, to which it was 

 subsequently traced. The last section was taken from Abrolhos 

 Island (lat. 26^ S.) to Tristan d'Acunha (lat. 26" S,), and thence 

 to the Cape of Good Hope, thus well illustrating the thermal 

 stratification of the South Atlantic, and clearly showing the 

 existence of a more voluminous cold underflow than existed in 

 the North Atlantic, probably owing to the fact that the Antarctic 

 circle was far less land-locked than the Arctic, so that there was 



