360 Wm. A. Norton on the Variations 



the positions of the maxima and minima. If fig. 25 be compared 

 with the curve of the mean daily variation of the horizontal 

 force, (see fig. 3, p. 40, in the July number of this Journal,) it 

 will be observed that the maxima and minima of the former fall 

 nearly midway between the maxima and minima of the latter. 

 Thus, the morning maximum of the horizontal force is at 5 

 a. m., and the minimum at 10 a. m. — and the morning mini- 

 mum of declination js from 7 to 8 a.m.: again, the evening 

 maximum of horizontal force is at about 3 p. m., and the princi- 

 pal maximum of declination is at 1 p. m., nearly midway be- 

 tween this and the morning minimum of horizontal force (at 10 

 a. m). The morning maximum of declination (at 2 a. m.) is 

 also nearly midway between the evening minimum of horizontal 

 force (about midnight), and the morning maximum of the same 

 (at 5 a. m). There is a small deviation from this general law in 

 the case of the evening minimum of declination, which occurs 

 some two hours later than the middle point of time between 

 the evening maximum and minimum of horizontal force. Be- 

 sides these relations between the maxima and minima of the two 

 curves, it may be seen that there are points of inflexion in the 

 curves of the horizontal force near the epochs of the maxima 

 and minima of the declination; and accordingly that when the 

 curve of the horizontal force is concave upwards the declination 

 (westerly) is increasing, and when it is convex upwards the dec- 

 lination is decreasing. These facts render it highly probable that 

 the diurnal variations of the horizontal force and declination are 

 linked together by some physical connection, as theory has al- 

 ready led us to suppose. 



Let us see whether this theory, besides suggesting the fact 

 of such a connection, can also explain the precise connection 

 which we have now found to subsist. If we recur to the prin- 

 ciples already laid down (p. 359), we shall see that the inquiry 

 before us leads us, in the first place, to seek for the daily 

 changes of position in the line of equal molecular magnetic force. 

 If we were to neglect the effect of dew and evaporation, the 

 line m question would be very nearly the true isothermal line 

 passing through the station of the needle. To simplify the 



matter we will for the present, consider the two as the same. 

 Now take some point (B), to the east of the station (A) of the 

 needle, situated on the isothermal line traced through A at 5 a. m., 

 about the time of minimum temperature : an hour later these two 

 points would not be on the same isothermal line, for the increase 

 of temperature at B would be greater than at A. (See curve of 

 daily variation of temperature.) The isothermal line through A 

 would therefore be directed to the north of R It is obvious that 

 this motion of the isothermal line toward the north, to the east 

 of the station A, will continue until the increment of tempera- 



