276 TRE ATLANTIC. [chap. v. 



there is a much larger body of water above them lieated by 

 conduction, convection, and mixture. 



The section between Montevideo and the meridian of Tris- 

 tan d'Acunha inchides, besides the soundings on the South- 

 American plateau and the soundings on the " cold wall," a se- 

 ries of soundings crossing the south-western trough with an 

 average depth of 2750 fathoms and an average bottom temper- 

 ature of — 0°*4 C, and a few soundings on the middle ridge 

 of the Atlantic, with an average depth of 1850 fathoms and 

 a mean bottom temperature of -|-1°'3 C, There seems to be 

 little doubt that in the trough a huge mass of Antarctic water, 

 at temperatures ranging from -)-l°*5 C. to — 0°-6 C, is creep- 

 ing northward at depths greater than 1800 fathoms. On the 

 central rise very little water at a temperature lower than -f l°-5 

 C. passes northward ; but that is only on account of the absence 

 of the required depth, for the isothermobaths of 1°*5 and 2° C. 

 are practically at the same levels respectively over the central 

 plateau and over the trough. But the evidence seems equally 

 cogent that the water at depths less than 1800 fathoms, and at 

 temperatures higher than 1°*5 C, is part of the same mass, and 

 is moving in the same direction. We can trace the same strata 

 continuously over the trough and over the eastern and north- 

 western basins, the temperature of each layer only very slightly 

 rising, as has been already shown, to the northward. 



Suppose a mass of water at a temperature gradually sinking 

 from the surface downward (Fig. 57) to be flowing slowly in 

 a certain direction, and suppose the course of that water to be 

 intercepted by a barrier which rises to the height of the layer 

 of water at a temperature of 2°'0 C. Suppose at the same time 

 that the water beyond the barrier is not constitutionally prone 

 to alter its temperature, and that it is quietly drawn off before 

 it has time to do so from any external cause. It seems clear 

 that the water beyond the barrier will be of the uniform tem- 

 perature to the bottom of the stratum of water which is passing 



