EARTH'S MAGNETISM TO THE SOLAR AND LUNAR PERIODS. 143 



17. In 1845 the maximum occm's between 14'' and IS*", or about 2'' after the 

 moon has passed the inferior meridian, the value of the vertical component then 

 diminishes till 21'' when there is a secondary minimum, it then increases consi- 

 derably till 2'" ; a secondary maximum occurring about lJ^ it again diminishes till 

 7\ when the principal minimum occurs, after which it increases rapidly to the 

 principal maximum. This year's result is little affected by the elimination of dis- 

 turbances ; the curve is rendered somewhat more regular. 



18. The results for the two years differ slightly from each other. The 

 principal maximum occurs about an hour later in 1844 than in 1845 ; the princi- 

 pal minimum occurs at the same time in both years ; but the secondary maximum 

 and minimum are not nearl}^ so well defined in 1844 as in 1845. The mean of 

 the two years gives the following periods : — 



The principal minimum about 5 hours before the moon's passage of the inferior meridian. 



The principal maximum about 3 hours after 



A secondary minimum about 4 hours before superior 



A secondary maximum about 1 hour after 



19. These periods are surprisingly near those indicated (No. I., 23), for the 

 horizontal component, the principal maximum and secondary minimum of the 

 vertical component occurring about an hour after those of the horizontal com- 

 ponent, while the principal minimum and sefcondary maximum occur an hour or 

 more before those of the horizontal component ; seeming to indicate that these 

 variations belong chiefly to the total intensity. 



20. As evidence of the accuracy of the results obtained after eliminating 

 disturbances, nothing perhaps could be more conclusive than the fact, that in 

 those months in which no disturbances occur, the general law is found weU 

 marked : it would not be difficult to bring this kind of evidence to bear, both 

 for the horizontal and vertical components. I shall satisfy myself at present by 

 giving the projection of one month's results, namely, the results for the lunation 

 included by December 15. 1845, and January 10. 1846 (Curves No. 3, Plate VI.). 

 From this single month the same, or very nearly the same, periods of maxima and 

 minima are obtained as in the mean for the two years. The maxima have nearly 

 equal values, and so also have the minima ; the principal maximum and minimum, 

 however, occur at the times of the secondary maximum and minimum of the mean 

 curve. Such a difference in the values of the maxima might be expected ; for, 

 though the moon has every declination in the course of one lunation, it is fuU 

 only at one of these declinations. It would not have been diflBcult to have ob- 

 tained a curve from one month's observations representing the mean curve better. 



21. In order to obtain the variations of the vertical component with respect to 

 the moon's age, summations were made similar to those indicated (No. I., 18) for 

 the horizontal component. It should be mentioned (as it should have been before 

 for the horizontal component), that in all the results for monthly periods mean 



