OCEAN TEMPERATURES 379 



parallel surfaces, varies from .008 to .02 for a wide range of tempera- 

 tures and salinities. But the idea of laminar flow no longer holds 

 when one considers the motion of whole volumes of water, hundreds 

 of meters in thickness, throughout which there is a pressure gradient, 

 as is often the case in oceanographic problems. In such cases one 

 should not use innere Reibung (viscosity), but Massenwiderstand 

 (hydraulic friction). 



The character of the motion is no longer simply laminar (one of 

 simple 'sliding to and fro, parallel to a given plane) but the particles 

 of fluid leave their surfaces and move in vortices along stream lines 

 transverse to the laminar motion. Thus a much greater resistance is 

 developed. For example, the values of the coefficient of viscosity 

 obtained by Nansen in his ocean researches are 200 to 40,000 times 

 the laboratory value. 



The same general idea has been successfully applied to certain 

 meteorological problems relative to wind, temperature, and humidity 

 by G. I. Taylor (1915). He found that the transfer, in a vertical 

 direction, of heat and water vapor in the atmosphere followed the law 

 of heat conductivity in solids, and that the effect of friction on the 

 motion of the air could be taken into account by substituting in the 

 general equations of motion a quantity called "eddy viscosity" for 

 the laboratory value of the coefficient of viscosity. 



From the observed relation of air temperature to height off the 

 coast of Labrador, he obtained values of the coefficient; of ''eddy 

 conductivity" from .57 X 10 3 to 3.4 X 10 3 , corresponding to wind 

 velocities varying from 2 to 3.4 Beaufort. Also from observations on 

 the relation of wind velocity and direction to height he obtained values 

 of the coefficient of eddy viscosity varying from .77 X 10 3 to 6.9 X 10 3 . 

 The values of these coefficients are more than 10,000 times the labora- 

 tory values, the ratio being of the same order of magnitude as that 

 obtained for sea water. 



In general, comparisons of his theoretical results with observations 

 indicated a very satisfactory agreement. An especially good agree- 

 ment was found between the predicted and the observed values of the 

 angle between the wind and the horizontal pressure gradient at 

 different levels. 



