t 5<) ) 



Fi'üiii Table II it appears that, in llie iieiglihourliood of ilic li^iil- 

 sliip llie veering of the wind in vviniertime is against sun, whereas 

 in other seasons it is with sun. Near de Hilt the veering is always 

 to the right as, in normal cases, onght to be the ease, owing to the 

 intkience of the earth's rotation. 



This abnormality is not accidental, dne to observational errors and 

 the smallness of the amj)litndes ; near the lightship Haaks and the 

 landstation Helder, the veering in wintertime is also to the left ; 

 near Helder not only during the wintermonths but in the autumn as well. 



The results for Helder are not given here because the friction 

 coetlicient, as deduced in the manner described above, does not 

 show reliable values in the different seasons ; evidently the series of 

 observations nsed for this purpose is of insuflicient duration. This 

 abnormal behaviour of the gradient, namely veering to the left, must 

 occur at every coast station where the sea is to the South and the land 

 to the North. If, however, the gradient veers against sun it is 

 quite possible that the wind still veers to the right for, if we put 

 i/j = 0, and thus assume that there is oidy an East — West gradient, 

 it appears from form. (5) that the wind veers with the sun ; the 

 possibility of a gradient veering to the left causing a wind veering 

 to the right is, therefore, not exclnded. 



In fact this case presents itself in summer time near de Bilt where 

 the gradient turns to the left whereas the wind veers to the right. 



At de Bilt also there is an important difference between spring 

 and summer on the one hand, winter and autunun on the other, as 

 shown in the ditferences of phase A ; in the former seasons this 

 phase difference is small and the ellipse approaches to a straight line, 

 in the latter the excentricity of the ellipse is much smaller. 



Tables V and VI, showing the elements of the wind- and gradient- 

 ellipses, convey a better idea of the results obtained and here they 

 can he easily corrected for the magnetical bearing by subtraction of 

 14'' from the angular \alues for Terschellingerbank. 



TABLE V. 



Windellipse of the monodiurnal variation. 



de Bilt. 



