417 



the same velocity of upwelling. Moreover, in applying hydrodynamical 

 equations to problems of oceanic circulation the coefficient of viscosity 

 must be replaced by another constant depending on the eddy, or 

 turbulent motion, and having a much greater value than the laboratory 

 value obtained from observations on a slow laminar flow free from 

 irregular motions. Similar results have also been found by G. I. 

 Taylor in certain recent studies of the temperature, water vapor, and 

 velocity in the atmosphere. The results of laboratory experiments 

 and theories based on them were helpful but could not provide the 

 numerical values required; in each case field observations were neces- 

 sary. Furthermore, since the eddy conductivity, or Mischungs- 

 intensitat, is not a physical constant of the substance, sea water or 

 air, but depends upon the intensity and character of the circulation, 

 its value will vary accordingly. The following approximate values of 

 these constants, the coefficient of viscosity, diffusion, and conductivity 

 under laboratory conditions and estimated from field observations in 

 the ocean and the atmosphere, illustrate the great differences between 

 field and laboratory conditions. 



TABLE 19 



Estimates of the coefficients of viscosity, diffusion, and heat conductivity made 



from field observations in the ocean and atmosphere compared 



ivith values obtained in laboratory experiments 



Sea Water 



Air 



1 The laboratory value of the "kinetic coefficient" of viscosity, or the coefficient of viscosity 

 divided by the density, is given since it is the constant in the equations of motion, which is 

 formally equivalent to the one given by field observations. 



2 The laboratory value of the thermometric conductivity is given since that is the constant 

 in the equation of heat conductivity, which is formally equivalent to the one given by field 

 observations. 



The estimation of the effect of upwelling on the surface tempera- 

 ture made necessary the consideration of results obtained for depths 



