"by using the equation 



P'U^ttH 



in the notation of Sverdrup-Munk, where ^ p is the wind pressure 

 against different portions of the wave, L the wave length, H the 

 wave height and U the wind velocity relative to the wave (the meas- 

 urements were taken with small wooden models of waves, placed in a 

 wind tunnel). With this notation equation (26a) per unit area (page 

 59) is written 



W^ = J p'sTr6(v - 6- )^ =IZI } [cm"-'- g sec'^] or [dyn cra"^], (26) 



taking (16) into account. Therefore 



p'(v - 6- rv6 ' 

 3 in our notation equals 2s in the notation of Sverdrup-Mtmk as given 

 above. The average value of s evaluated by these authors from Stanton's 

 measurements is s = 0.049. 



The effective wind stress is given by (24) as a function of 

 wind velocity and C(p), where the latter value takes into account 

 the different "friction conditions" of the rough sea surface depend- 

 ing on the stage of wave development. Because p = p in the fully 

 arisen sea only depends upon the wind velocity v (see formula (60)), 

 the frlctional coefficient C(p ) is given as a function of v, too. 

 By this reason a quadratic relation between the effective stress 

 and the wind velocity v is not to be expected at the sea surface with 

 its changing roughness conditions, and in fact not observed at all 

 [13]. When the assumption is made that the vertical distribution 

 of the wind velocity in the lower layer of the atmosphere immediately 

 over the sea surface may be represented by a logarithmic law, then 



65 



