from the Falkland Islands to Monte Video. 631 



already come to an end, between long. 28°-9 "W. and the coast of 

 America. 



The distribution of the lower temperatures is much the same at station 

 131 as it is at station 327, the isothermobaths sinking slightly. At 

 station 129, which is right in the middle of what may be supposed to be 

 the direction of the current, the depth is 2150 fathoms, and the tempe- 

 rature 0*6 C. ; the cold-water trough has therefore probably stopped 

 before reaching this point ; all the belts of water at given low tempe- 

 ratures are evidently continuous, although the isothermobaths of 3° C. 

 and 2°*5 0. are now sinking rapidly. The accumulation of water above 

 3° 0. in temperature is still more apparent at Station 112, nearly under 

 the equator. 



The next station, INFo. 15, is the first of a series of rather deep sound- 

 ings taken after crossing the "Dolphin Rise " (the central bank of the 

 Atlantic) on our voyage from Teneriffe to Sombrero in 1873. The 

 deepest of these soundings are a little above 3000 fathoms ; and they 

 look from their position very much as if they were in the axis of the 

 Antarctic indraught ; but the temperature in no case falls below l°-3 O. ; 

 and, tracing southwards, we find that a belt of shallow soundings (not 

 more than 2000 fathoms) extends from the central plateau about lat. 

 21° N., long. 46° 30' W., to the coast of South America, somewhere 

 about Cape Orange. Going northwards, we encounter much greater 

 depths, but no temperatures lower than 1 0, 3 C. ; the whole mass of water 

 to the very bottom rises steadily, though only slightly, to the northward, 

 the merely local depression, which never reaches l°-3 C. along the line of 

 the Labrador current, being scarcely worth mention. It is evident that 

 the water of the Atlantic is not generally affected by an indraught from 

 the Arctic Sea. The diagram (Plate 33) shows very well the effect on the 

 relative positions of the isothermobathic lines of the rapid removal of 

 the warm surface-water over the region of the equatorial current and 

 counter- current in the gathering up of the higher lines near the equator, 

 and of the curl of the warm water to the north and south as the Grulf- 

 stream and the Brazil current after the bifurcation of the equatorial 

 current, in the depression of the isothermobaths of 20°, 15°, and 10° O. 

 to the north and south of the equatorial region. 



The causes of the very perceptible rise in temperature of the whole 

 mass of the water of the North Atlantic will require further investiga- 

 tion. As I have said elsewhere, I am inclined to attribute it mainly to 

 the banking in and accumulation of the northern reflux of the equatorial 

 current, and its gradual mixture with the water beneath ; and it seems 

 to me that this cause is adequate to the effect. Still, even if this be so, 

 we have to discover how that mixture takes place ; and as there appears 

 reason to believe that evaporation is in excess of precipitation in the 

 Atlantic, it seems almost certain that differences in specific gravity and 



