heights, between 4 and 200 cm above the wave crests, using cup anemometers. 

 At the same time the following wave characteristics were measured; wave 

 direction; wave height, h; wave period, T; and wave velocity, C (from 

 which the wave length, \ , could be determined by use of the equation 



\ = CT). In addition, the depth of water and the tidal motion (direction 

 and velocity) were measured. Temperature and humidity were also closely 

 noted. Details of the measurements have been published elsewhere^ 14 ). 

 These investigations made during 1947 (will be continued in 1948), have 

 resulted in a total of 193 wind velocity-profiles (simultaneous wind 

 measurements for 6 different elevations) with wave and current measurements, 

 and 73 wind velocity profiles over the dry tidal flat. 



In addition, special measurements were made of such things as wind 

 profiles from 2 to 24 cm over the ripples of small shallow pools, such 

 as remain at low tide on the otherwise dry tidal flats, and three 

 dimensional wind profile measurements to investigate the change of wind 

 profile with the change of base (land-sea, land-tidal flat). 



Before evaluating these data the influence of tides was eliminated 



by plotting all wind velocity measurements on a coordinate system relative 



to the water surface. The individual tide-corrected wind profiles were 



then summed up into average wind profiles in order to reduce the influence 



of scatter. These were, without exception, in conformity with the 



logarighmic formulas and from these, the respective values for roughness, 



z , and shear stress velocity, u*, were determined using the method of 



«o , 



least squares. 



Since we wished to establish whether the proportionality between 

 roughness and wave height for sharp crested waves as found by Motzfeld 

 was also valid for traveling sea waves, wave heights were first used as 

 the organizing principle for summarizing individual runs into average 

 values. Since the depth of water varied between 1 and 175 cm in our 

 measurements, the wave heights also covered a relatively wide band from 

 0.1 to 30 cm. It was necessary to apply a correction to the elevations 

 of the wind gages, as the datum had been the plane of the crest of waves. 

 A new datum, the mean water level, was arrived at by adding a half of the 

 average wave height. The determination of the roughness parameter for 

 the various ranges of wave heights revealed no relationship between 

 roughness and wave height. The relation(2) found by Motzfeld with models 

 of sharp crested waves did not seem to be substantiated, at least for 

 shallow water waves. 



Next an investigation was made as to whether the distribution of 



wind velocities^), as derived from model experiments by Motzfeld on waves 



with round crests, were applicable for sea waves. For this the individual 



wind profiles had to be arranged according to shear stress velocities, u*, 



(that is, practically according to the wind velocities) and within these 



limits according to wave form (tan a ). 



m 



The correct value for the maximum slope of the wave profile could 

 only be derived if the exact slope of the wave were known. This, however, 



