Vertical Changes 



137 



tion from one water layer to another is negligible. Conduction 

 through quiet water is similarly very slow. If all the water in Lake 

 Constance with a depth of 100 m were cooled to 0°C and then the 

 uppermost layer heated to 30°C, more than 100 years would be re- 

 quired for a measurable amount of heat to reach the bottom by radia- 

 tion and conduction alone. It is obvious that heat can be transferred 



Temperature, °C 

 12 16 



20 



24 



28 



> Deep layer 



20001- 



Modified from Iselin, 1936 

 Fig. 5.4. Permanent vertical temperature structure in the deep water of the 

 central North Atlantic. The seasonal changes in upper layers are indicated: (A) 



April, (B) August, (C) December, (D) February. 



to lower levels only by the actual movement of the water, that is, by 

 vertical circulation. 



We are further convinced of this conclusion when we examine situa- 

 tions where vertical circulation has been prevented. When the tem- 

 perature of the bottom water of a pond has been reduced to 4°C and 



