632 Prof. W. Thomson on the Voyage of the ' Challenger ' 



other causes come into play, though not very obviously. It is most 

 probably a phenomenon of considerable complexity. 



I will not refer to the thermal conditions of the eastern trough of the 



o 



Atlantic at present j they are somewhat different, and will require 

 separate illustrations after we have recrossed the basin, which we hope 

 to do in the next six weeks. 



To recapitulate the general facts and conclusions with regard to the 

 distribution of ocean temperatures in the western trough of the Atlantic, 

 it seems to me : — 



1. That this trough (and the Atlantic as a whole) must be regarded in 

 the light of an inlet or gulf of the general ocean of the " water hemi- 

 sphere," openiug directly from the " Southern Sea." 



2. That the water of the Southern Sea simply wells up into the 

 Atlantic, and that all the temperature bands of the Western Atlantic are 

 essentially continuous with like temperature bands in the Southern Sea, 

 with these modifications : — that (a) above a certain line, which may be 

 roughly represented by the isothermobathic lines of 5° and 4° C, the 

 temperature of the water is manifestly affected by direct radiation and 

 by the very complicated effects, direct and indirect, of wind-currents ; 

 and (6) that the whole mass of under-water gradually and uniformly 

 rises in temperature towards the head of the gulf. 



3. That water at any given temperature (below 4° C.) can only occur 

 in the Atlantic where there is a direct communication with the zone of 

 water at the same temperature in the Southern Sea without the inter- 

 vention of any continuous barrier. (The actual result of the present 

 arrangement of such barriers is, that, however great the depth may be, 

 no water at a temperature lower than 1 0, 3 is found to the north of the 

 equator, and no water below the freezing-point in any part of the 

 trough except in the depression already described between the mouth of 

 the River Plate and Tristan d'Acunha.) 



4. That the water of the Atlantic is not sensibly affected in tempe- 

 rature by any cold indraught from the Arctic Sea. (I purposely neglect 

 the Labrador current and the small branch of the Spitzbergen current ; 

 for these certainly do not sensibly affect the general temperatures of the 

 North Atlantic.) 



5. That although there is a considerable flow of surface-water through 

 the influence of wind-currents from the Atlantic into the Southern Sea, that 

 flow does not seem to be sufficient to balance the influx into the basin of 

 the Atlantic (the constant influx being proved by the maintenance of a 

 general uniformity in the course of the isothermobathic lines, and by the 

 maintenance in all the secondary basins of the minimum temperature due 

 to the height of their respective barriers) ; that for several reasons (the 

 lower barometric pressure and the supposed greater amount of rainfall in 

 the Southern Sea, the higher specific gravity at the surface than at greater 



