220 



ed that these motions in water bring about the formation of the mixed surface 

 layers which often extend many meters in depth. In the air it seems probable 

 that these motions are important in forming the layer of nearly constant potential 

 temperature and mixing ratio which commonly extends in some regions of the 

 North Atlantic from the sea surface up almost to cloud base. Large sea-salt 

 particles, originating in "white caps", and having settling rates of 1 to 20 centi- 

 meters per second are also found well distributed in this sub-cloud layer. A 

 vigorous overturning of this layer of air is thus indicated. 



Observations such as those briefly indicated above give a qualitative indi- 

 cation of the form and some of the rates of the fluid motions in the air- sea bound- 

 ary region. These motions are probably a result of the transfer of properties 

 between the sea and the atmosphere. Instruments, capable of measuring the 

 various properties of the fluids which are being continuously removed upward and 

 downward from the air-sea interface, should contribute largely to our under- 

 standing of this important boundary. It is my opinion that new instruments are 

 required which are realistically designed to make measurements and to obtain 

 samples within specific and rather limited areas of the fields of fluid motion in- 

 dicated. In the sea, measurements of the temperature, salinity, velocity, ox- 

 ygen and perhaps plankton content of descending waters should be made under the 

 areas of convergence. In the air, measurements of temperature, water vapor, 

 velocity and sea salt content in the ascending regions are needed. Measurements 

 such as these seem essential in order to test the reality of the motions apparently 

 delineated by the sea birds and the Sargassum, and in evaluating their possible 

 role in the transfer of properties between the sea and the air. 



DISCUSSION: R.B. Montgomery 



It may be of interest to supplement Dr. Munk's paper with an account of 

 the present position of the study of evaporation from the ocean, especially since 

 this study cannot proceed without the development of suitable instrumentation. 

 What is presented about evaporation has application to (sensible) heat transfer 

 also, as the transfers of momentum, water, and heat are closely related, but 

 heat transfer is usually less important and more complicated. 



During the 1920' s and early 1930's, an effort to learn about evaporation 

 from the ocean was made by maintaining evaporation pans in gimbals on research 

 ships (e.g., "Carnegie," "Meteor," "Atlantis"). The observations so obtained 

 have not served any purpose. It is now realized that pan evaporation does not 

 determine the evaporation from natural surfaces, whether water, ground, or 

 vegetation. 



A vvay to measure evaporation from the ocean is not yet at hand, but in my 

 opinion we are now in a position to see how to develop a satisfactory method. In 

 a preliminary form, this method has already been used by Jacobs (1942) to calcu- 

 late climatological evaporation from climatological values of winds and tempera- 

 tures over the ocean, but so far we have no reliable check on the constants he 

 used and on other details. 



This method depends on the apparent fact that unit- area rate of evaporation, 

 E, is primarily a function of two quantities, the specific-humidity difference bet- 

 ween the sea surface and some suitable height, Aq, and the wind speed at some 

 suitable height, u, 



E = f(Aq, u). 

 The function is more or less of the simple form 



