100 MR BROUN ON THE RELATION OF THE VARIATIONS OF THE 



maximuni is attained. It now commences to decrease slowly till 8 p.m., more 

 rapidly from 8 till i), causing another inflexion in the curve, slowly again till 

 2h 20™ A.M., when there is a minimum; the force then increases slightly till 

 6" 30"" or 40"" when a maximum occurs, after which it diminishes rapidly till 

 10'' 20" A.M., the period of the minimum. These hours differ somewhat fi-om 

 the periods obtained at other observatoi-ies ; and while some part of these differ- 

 ences may be due to errors of temperature correction, I do not think that such 

 errors wUl altogether account for them, but that the accurate periods of maxima 

 and minima will be found to differ at different places. At Toronto in Canada, 

 for example, the maximum occurred a little after 4 p.m. in 1842 ; and as the 

 mean temperatiu-e of the magnet at the succeeding observation hour differs but 

 little from that at 4% the period cannot be affected by temperatm-e. Some 

 observatories shew the maximum as late as 7 p.m. It does not, however, seem 

 improbable, that the periods of maxima and minima should differ at differ- 

 ent places, when it is known that these periods vary at the same place in the 

 com'se of the year ; at Makerstoun, in 1844, the afternoon maximmn occurred as 

 early as 3^ 10™ in December and January, and as late as C" 50" in June ; the 

 minimum at 10'' 20" a.m. in the winter mouths, and at 9" 40™ a.m. in June ; the a.m. 

 maximum occurs at 6" 40"" in December, and about 5'" in the summer months, 

 while the eai'liest minimum occurs nearer midnight in winter than in summer. 

 In this way the periods of the principal maximimi and minimum approach to 

 each other, and to noon in winter, and remove from each other, and from noon in 

 summer. (See Curve, No. 1.) The reverse to some extent takes place with regard 

 to the periods of the secondary maximum and minimum, which remove fi-om each 

 other in winter, and approach each other in summer, tUl in June the maximum 

 and minimum seem to destroy each other. 



The morning maximum is greater than the afternoon one in December ; — ^in 

 November, January, and February, they differ but little from each other ; and in 

 December, January, and February, the two minima are nearly equal. 



4. The inflexions noted in the mean curve about 3 p.m. and 9 p.m., become 

 minima in the winter months, so that there are then three or four maxima and 

 minima ; the smaller ones nearly compensate each other in the mean of the winter 

 months, as they occur at different hours in each month.* I shall consider the cause 

 of these secondary afternoon maxima and minima on another occasion. With 

 regard to the 2 a.m. minimum and 6 a.m. maximum, these seem nearl}^ to vanish 

 in the summer months. In the means for the months of June and July they can- 

 not be detected, excepting that the intensity decreases more rapidly after 6 a.m. 

 than before it ; it should not be concluded on this account that this maximum and 

 minimum do not exist. Having projected the hourly observations made in each 

 day of June and July, I have not found one day in ten on which the secondary maxi- 

 * November, December, and January, have been taken as the three winter mouths. 



