352 



SCIENCE PROGRESS 



all stations : for a rough approximation we may take the datum 

 level as sea level and take the velocity at any point above our 

 heads to be proportional to the surface wind and to the height 

 of the point above sea level, so that the facts may be roughly 

 represented by fig. 3. 



Ox represents sea level, points along Og heights above sea 

 level ; h is the height of the surface anemometer above the sea 

 level, H the height at which we want to find the velocity. Take 

 hv to represent the observed wind and draw HV parallel to 

 Ox. Join Ov and produce it until it meets HV in V. Then HV 

 is the effective velocity of the wind at the height H. 



The increase of wind thus described will go on until the 

 gradient velocity is reached. Take OG to represent the gradient 



o G 



Fig. 3. — Variation of wind with height above sea level in the surface layer. 



velocity. Set out GM, a vertical in the diagram. Produce OvV 

 to meet GM in M. Then the proportionate increase will go on 

 until the point M is reached. 



This is a rough-and-ready representation of the phenomena 

 of the surface layer. I am aware that objections can be raised 

 to it on practical and theoretical grounds but I will not now 

 discuss the evidence for its practical effectiveness nor attempt a 

 better theoretical curve than the straight line to represent the 

 transition from the surface wind to the gradient wind ; I would 

 recommend aeronauts to use this rough approximation and see 

 how it works. For many purposes it is sufficiently near the 

 truth. It should, however, be remarked that the regular in- 

 crease of velocity up to the gradient-wind is deduced from 

 observations made by means of kites and pilot balloons. The 

 observations all belong to the daytime and the inference is only 



