Turning of the Wind. 109 



occurrences of the different winds, if we adopt the number 

 of registrations of eastern winds as unite : — 



S 4-3 Between N.E. and K. . 55 



Between S. and S.E.. 3-1 N 5-5 



S.E 2-4 Between N. and N.W. 3*5 



Between S.E. and E. 1-6 N.W 26 



E 1-0 Between N.W. and W. 2*2 



Between E. and N.E. 1-9 W 33 



N.E 4-3 Between W. and S.W. 3"7 



SW 2-7 



S.W. and S... 3-8 



Consequently, by giving the means we have — 



Between S.W. and S.E 326 



S.E. and N.E 2-24 



N.E. and N.W 4.28 



N.W. and S.W 270 



We readily perceive that the celerity with which the vane 

 travels through each quadrant is in an inverted ratio with 

 the frequency, and consequently that between S.E. and 

 N.E., N.W. and S.W., the vane travels more easily than 

 between S.W. and S.E., N.E. and N.W. The two latter 

 quadrants form the contesting fields. Before leaving the 

 direct proofs, as derived from yearly observations, I should 

 like to add a few words having reference to the mean direc- 

 tions of the wind in this country. 



In computing this quantity for the year, the formula given 

 in statics for finding the mean direction, when several powers, 

 acting in one and the same plane, attack a point, has been used. 



The force of the wind was taken as equal to one, and the 

 relative frequency as the components. By this method I 

 obtained the mean direction of the wind for the year to be 

 N. 38° 54' W. 



Taking the relative force into consideration, these results 

 will be slightly changed, as I shall show at another oppor- 

 tunity. I have only endeavored to prove the greater predomi- 

 nance of the " Equatorial " over the " Polar Current." 



Following the mean direction of the wind through the 

 different quarters, we find — 



For the quarter March to May N. 50° 45' E. 



„ „ June to August . . N. 14° 31' W. 



„ „ September to Nov. S. 80° 20' W. 



December to Feb.. . S. 17° 44' W. 



