( 53 ) 



42°, tlie great axis of the gradient cllipso is directed towards (N — S) ; 

 for a latitude of 42° //, =: //.^, and the ellipse becomes a circle; 

 more southerly the great axis points towards (E — W) ; at the equator 

 //j = and the ellipse is flattened down to a straight line directed 

 E— W. 



If E — as is probable — is not exactly proportionate to the 

 third power of .sm(f, the limit 42° varies of course accordingly. 



4. In two recent publications Gold ^) has compared this theory 

 with the results of the observations; for this purpose he chooses the 

 following method : the gradient variation, as deduced from the semi- 

 diurnal variation of the atmospheric pressui'e is assumed to be known ; 

 from form. (1) or, what comes to the same, form. (4) the four 

 wind constants A, B, C\ and C, are calculated and the results arc 

 compared with the wind variation, as deduced IVom the observations 

 by giving k the values 0, ^ and 1. 



Obviously this method leads to very complicated and almost 

 unmanageable expressions for the wind \ariation and the inverse 

 way, namely taking the wind variation as a known quantity and 

 then calculating the barometric variation, would be a simpler and 

 equally suitable method. 



Calculating Gold's formulae for a latitude of 52°, the average 

 latitude of the Netherlands, we tind, 



V It 



k = 45 cos (2 nt — 33°) 44 cos (2 nt — 1 23°) 



/.• =r 1 28 cos ( -' nt — 343°) 27 cos (2 nt — 72°) 



k= ^ 17 cos (2 nt — 328°) 16» cos (2 ?ii — 55°), 



In calculating these expressions, the semidiurnal variation of 

 atmospheric pressure as found for de Bilt has been use J; the 

 amplitudes are expressed in cm. p. sec. 



On the average, taken over the whole year, the result of wind 

 observations made at two Dutch stations is : 



V u 



de Bilt 16.8 cos (2 nt — 333°) 1 7.4 my (2 nt — 65°) 



Terschellinger bank 1 1 .1 cos (2 7it — 334°) J 1 .9 cos (2 nt — m°) 



The agreement between theory and observation is very satis- 

 factory, much better than for the two cases as calculated by Gold, 



^) E. Gold. The relation between periodic variations oC pressure, lemperature 

 and wind in the atmosphere. Phil Mag. 1909, p. 26—109. 



Id. Note on the connexion l)elween the periodic variations of windvelocily and 

 of atmospheric pressure. Publ. Meteor. Olfice, 203, 1910. 



