of Terrestrial Magnetism. 435 



the bifilar magnetometer is similar to that of the declinometer, 

 but with a retardation of three hours. 



Explanation. — A glance at Colonel Sabine's figures in the 

 second volume of the Hobarton Observations, plate 1, p. 5, for 

 the declination, and at plate 4, p. 43, for the component of the 

 bifilar magnetometer, will be sufficient to show, that while the 

 minimum of the declination occurs between 20 h and 21 h , and 

 the maximum at about 2 h , the minimum of the bifilar magneto- 

 meter occurs about 23 h , and the maximum between 4 h and 5 h . 

 See also the figures in which this celebrated author makes the 

 comparison between Hobarton and Toronto in the first volume 

 of the Hobarton Observations. At p. 34, plate 1, the curves of 

 the declination are shown; and at p. 54, plate 2, figs. 1 and 2, 

 that of the horizontal force; the perfect agreement of the 

 curves will be seen, (though the scales of the abscissas are differ- 

 ent) and the same retardation between Hobarton and Toronto 

 which has been already remarked in the declinations. 



Next let us consider the peculiarities of the equatorial obser- 

 vatories. At St. Helena a singular law holds in the horizontal 

 force. It has a single simple period, and the only indication of 

 a secondary period is that the axis of the abscissae is not divided 

 by the curve into equal parts, but the diurnal part is less extended 

 than the nocturnal. See the St. Helena Observations, p. 30, 

 plate 4, fig. 3. 



Here, then, the semi-diurnal period vanishes entirely or nearly 

 so, and cannot be compared with that of the declination. 



But we shall presently see, in the theoretical law, the true ex- 

 planation of this singular fact, and it will be one of the principal 

 proofs of the theory which we are about to expound. In the 

 variation of this component two very decided periods are evident, 

 the diurnal and the annual. The diurnal maximum occurs 

 between 23 h and h , and the minimum at 9 h or 10 h ; but in 

 May and June (the winter months) it occurs later, viz. at ll h . 

 The form of the curve shows a rapid increase and an equally 

 rapid decrease. The annual variation is likewise remarkable for 

 its simplicity, being a perfect curve of sines. See p. 28, plate 2, 

 fig. 2. We shall see further on, that at Bombay (lat. 18° 53' N.) 

 the phases approximate to those of St. Helena, except that the 

 secondary periods are more sensible, though not much more so. 

 From the two extreme cases which we have here considered, we 

 may infer what would happen at an intermediate station, like the 

 Cape of Good Hope ; that there would be a diurnal and semi- 

 diurnal pel iod, but the latter considerably less developed than at 

 Hobarton and Toronto. Observation confirms this ; and fig. 1, 

 p. 40 of the Cape of Good Hope volume shows it at a glance. 



The period, then, of the horizontal force is the resultant of 

 2F2 



