M. A. Secchi on the Periodical Variations 



pended from them, would remain in the plane of the meridian ; 

 but a magnetized bar being suspended, their line of suspension 

 undergoes a torsion until the bar takes a direction at right 

 angles to the magnetic meridian. The force of torsion exerted 

 by the threads counteracts the magnetism of the earth, and the 

 needle is rendered in a manner astatic, so as to change its posi- 

 tion with the slightest variation of the terrestrial force. Obser- 

 vations with this instrument are made, with the same ease as 

 those of the declination, by -means of a mirror or collimator 

 attached to the magnet, and observed by a telescope with a gra- 

 duated scale. One precaution must be taken ; as the force of 

 the magnet varies in intensity with the temperature, it is neces- 

 sary to take account of this, so as to correct the observations for 

 its influence by known methods of sufficient accuracy. 



The vertical force is measured by means of the balance mag- 

 netometer, invented by Dr. Lloyd. This consists of a bar sup- 

 ported horizontally on knife-edges, which rest upon two agate 

 planes, exactly like the beam of an ordinary balance. This ia 

 arranged in the plane perpendicular to the magnetic meridian, 

 and is adjusted by means of small weights, so as to undergo a 

 sensible inclination with the slightest change of the vertical 

 magnetic force. Its motions are observed by means of micro- 

 scopes attached to the stand of the instrument, and collimating 

 on the intersection of spiders' webs fixed at the ends of the bar, 

 or by observing, as before, the image of a scale reflected by a 

 mirror attached to the bar. 



When the variations of the horizontal and vertical force are 

 obtained, the variations of the inclination are deduced from them 

 by known formulae. The balance magnetometer, as well as the 

 bifilar, is subject to variations depending on the temperature, 

 and therefore it is necessary to determine the coefficients of the 

 correction. It does not, however, appear that its results are in 

 all cases as certain or as exact as those of the bifilar. 



A. Horizontal Force. 



The component which we are now considering is that which is 

 obtained from the bifilar magnetometer, arranged at right angles 

 to the magnetic meridian. The variations may be expressed in 

 the following manner. 



General Laws. — The bifilar magnetometer is subject to a 

 horary variation of a double period, diurnal and semi-diurnal ; 

 in the semi-diurnal period the magnitude of the variation depends 

 on the geographical latitude, and is zero at the equator; the 

 phase depends on the angle which the sun makes with the mag- 

 netic meridian. 



We will demonstrate this by steps. Beginning with the sta- 

 tions of middle latitude, there is this simple law ; the curve of 



