178 PHYSIOGRAPHY. [chap. 



was very carefully examined by means of instruments 

 specially constructed to avoid sources of error. These ob- 

 servations show that, as a rule, the temperature diminishes 

 as you descend, just as was shown to be the case in the 

 North Atlantic. Reference to Fig. 45 shows that the 

 bottom-water of that part of the ocean has a temperature 

 only a little above 30° F. ; while, in other places, it is still 

 lower, and may even descend below the freezing-point of 

 fresh water.^ It appears that the presence of such cold 

 water in the deeper parts of the ocean, even in tropical 

 regions, can hardly be explained otherwise than by assum- 

 ing a grand movement of water from the polar towards the 

 equatorial regions. Dr. Carpenter has brought forward 

 much evidence to prove the existence of such a general 

 oceanic circulation, and he refers the movement mainly tot 

 differences of density due to differences of temperature. 

 The cold polar waters sink by their density and form a 

 deep layer, which creeps along the ocean-floor towards the 

 equatorial regions ; while the warmer and relatively lighter 

 water floats on the surface in a contrary direction, or from 

 equatorial towards polar areas. ^ By such means, a complete 



^ The freezing-point of water is lowered by the addition of common 

 salt, and ordinary sea- water does not freeze until reduced to 28° '4 F. 



^ The character of this circulation will of course be greatly modified 

 by the shape and depth of the sea-bottom over which the cold water 

 creeps. Thus, the southern part of the Atlantic basin communicates 

 freely with the Antarctic Sea, and the influx of cold water is there- 

 fore unimpeded ; but the nortliern ^rt of the basin is contracted, and 

 the principal channel through which Arctic water can flow southwards 

 is the shallow channel between Greenland and Iceland ; hence the 

 underflow of glacial water from the North vinll be much less than that 

 from the South. This is still more strikingly shown by the shape of 

 the great Pacific basin, where the communication with the Northern 

 Polar seas is confined to the narrow and shallow channel of Behring's 

 Strait, through which very little glacial water can flow to the south. 



