74 



DEACON AND WEBB 



[chap. 3 



temperature, then various samples of air that result from mixing will be 

 represented by various positions along tbe straight line joining these two 

 points. On account of the meagre quantity of data re])resented by the plots in 

 Fig. 15, they can do little more than merely illustrate the principle involved; 

 and, clearly, it would have been advantageous to make observations over an 



10 II 12 13 14 15 16 



Potential vapour pressure, mb 



(a) Strong wind U|3 greater than 7 m/sec 



9 10 II 12 13 14 15 



Potential vopour pressure, mb 



(b) Moderate wind U|3 between 4-5 m/sec 



9 10 II 12 13 14 



Potential vapour pressure, mb 



(c) Ligtit wind: u^-^ less than 3 m/sec 



Fig. 15. Temperature and humidity at 12.75 and 4 m and temperature at the sea surface, 

 plotted on the characteristic diagram — observations (15-min averages) in Port 

 Phillip Bay, May, 1958. The saturation curve for air in contact with sea- water is 

 shown. Symbols identify different runs. 



extended range of heights. However, we can at least note the following conclu- 

 sion. In strong and moderate winds, the extrapolations of the atmospheric 

 observations meet the saturation curve close to the measured sea temperature — - 

 within 0.5°F in every case. In light winds, the extrapolations point generally 

 below the measured sea temperature, the disparity being greater than 0.5°F in 

 nearly every case : this is suggestive of the existence of a cool-water skin layer 

 under the favourable circumstance of a smooth surface. 



