shearing force) is greater than the component of pressure. This conclusion 

 is confirmed by the coefficient of pressure and friction resistances, c 

 and c , as determined by Motzfeld. For wave profiles with sharp crests? 

 the coefficient c of the pressure resistance amounts to approximately 87 

 to 88 percent of the total coefficient of resistance. The effect of 

 pressure for wave profiles with round crests decreases considerably in 

 comparison to the effect of friction. Thus the coefficient c . of the 

 pressure resistance for sine, or as the case may be, trochoid profiles, 

 with a ratio of wave height/wave length = 0.1 (steep profiles) amounts to 

 39 to 44 percent of the total value, and reduces finally to 18 percent for 

 a profile corresponding to a sea wave with regard to steepness (wave 

 height/wave length = 0.05). From the experiments on models made by Motz- 

 feld it seems to follow that for waves with sharp crests the pressure re- 

 sistance is the controlling factor, while for round wave profiles, the 

 friction resistance is the controlling factor. 



We must now establish whether and to what extent the knowledge ob- 

 tained by Motzfeld from his experiments with models holds for air flow 

 over moving sea waves. Prior to this time there have been no suitable 

 observations available. In the published measurements of wind velocities 

 (Wust^ 8) , Montgomery^ 9 ), Shoulejkin (10) and BruchClD) as well as in the 

 wind gradient measurements obtained during the International Gulf Stream 

 Investigation in 1938 aboard the ALTAIR, the data were taken within a 

 few meters immediately above the water surface and wave measurements were 

 not made simultaneously; therefore, the relationships found by Motzfeld 

 for his wave models could not be accurately checked. This explains why 

 Rossby(9) found a tendency for the roughness of the sea surface to decrease 

 with increasing wind velocity, while Sverdrup(*2) postulated an increase, 

 and ModeldS) derived a constant from the Bruch measurements. The 

 difficulty seemed to arise from the definition of the roughness parameter 



as z . 

 o 



The author, under the sponsorship of the Maritime Division of the 

 Office of Meteorology for North West Germany, Hamburg, has, among other 

 things, made measurements of wind velocities in the air layer immediately 

 adjacent to the water surface simultaneously with wave measurements. 

 Because of lack of time and the difficulties inherent in such measurements 

 over deep water, the investigations were undertaken over shallow water. 

 Thus, there existed the possibility of mounting measuring instruments to 

 a mast rigidly implanted in the bottom of the sea. These measurements were 

 made in the tidal flats northwest of the island of Neuwerk during July 

 and October 1947. 



The advantages of this location, aside from the mentioned possibility 

 for erection of the mast, consisted of a uniform depth of water of some 

 175 to 200 cm at high tide over a large area, and the relatively great 

 regularity and thereby good measurability of the waves. A disadvantage 

 was the considerable tidal motion with its associated variation in water 

 depth. 



The average wind velocity was measured simultaneously at 6 different 



