Chase.] 



160 



[October. 



with the force which keeps the planets in their orbits appears to be 

 complete and conclusive. 



It is possible that a careful study of the relation of the winds to 

 the various magnetic variations would bring to light other evidences 

 of parallelism as striking as the one I have already pointed out 

 (Proc. A. P. S., X, 104) between the curves of vertical magnetic 

 force and force of wind. Such a study might require special atten- 

 tion to the pressure and velocity of the wind, the times of maxima 

 and minima, and other particulars, the need of which would be sug- 

 gested by experience. 



Whatever cause affects at the same instant the magnetic and aerial 

 currents should first manifest itself through its influence upon the 

 needle, on account of the amount of inertia in the air. Upon ex- 

 amining the second volume of the St. Helena Observations, which 

 contains a record of the direction of the wind at intervals of sis 

 hours, so arranged as to facilitate a comparison with the magnetic 

 declination, I find in each year, from 1844 to 1847 inclusive, that at 

 one hour before the observation of the wind (and in each year except 

 1846, at the hour of observation) there was a greater average westerly 

 declination when the wind was nearly east than when it was nearly 

 S. by E. This is shown by the following table, which embraces all 

 the tabulated instances when the wind was E. by N., E., E. by S., 

 or S. S. E., S. by E., or S. 



The variation of declination is ascertained, 1st, by subtracting 

 from the observed declination the monthly mean at the same hour; 

 2d, subtracting the monthly average of the daily means from the 

 mean variation of the day; and 3d, subtracting the latter result 

 from the former. One scale division of the declinometer =:z O'.Tll- 

 Increasing numbers denote decreasing westerly declination. 



* January to July inclusive. 



