244 Mr. Michael Donovan on certain Improvementfi in 



be in its condition of greatest sensibility, we must proceed as follows. It is to 

 be transferred to the volta-magnetometer, and its spindle or common axis is to 

 be confined between the blades of the forceps. The upper bar is to be turned 

 on the common axis until it form any angle with the lower one, suppose 80°: 

 the forceps being opened, the needle will oscillate, and finally settle in such a 

 position, as will show the relative intensity of the poles by tlie degrees pointed 

 to on the graduated circles. If the magnetism be equal, the four poles will 

 stand at 40^ But it has been already shown, that this equality may be but 

 apparent, owing to the want of coincidence between the magnetic and metrical 

 axes. In order to test this, the bars are to be closed imtil they fall into the same 

 vertical plane, the poles being reversed. If the compound needle, when libe- 

 rated, after oscillating a while, turn very slowly into the magnetic meridian, no 

 more need be done ; for although some httle irregularity in the distribution of 

 the magnetism is thus manifested, the desired efiect is obtained. If the needle 

 do not turn into the magnetic meridian, the bars are to be again opened to an 

 angle of 80^ one of the poles is to be touched with the opposite pole of a strongly 

 macrnetized steel wire, or sewing needle, until there be a difference of 1° on the 

 indications of the bars. This, or less, will be sufficient difference, when the 

 bars are closed, to carry the compound needle into the magnetic meridian. 

 Should the difference be more than 1°, that bar nearer the magnetic meridian 

 line should be touched with the similar pole of the magnetized wire, and by 

 lessening or increasing its power, the difference of 1^ may be attained. The 

 spindle being then caught in the forceps, the upper bar of the needle is to be 

 turned round on the common axis, until the reversed poles appear in the same 

 vertical plane. But a want of precise coincidence in this respect, between the 

 reversed poles of the bars, to an amount often undiscoverable by the eye, will 

 cause the compound needle to lie at right angles with the magnetic meridian. 

 Hence, to produce perfect coincidence, the eye must be assisted by a magnifier, 

 and the compound needle must be executed with great precision so that the 

 bars shall be straight, and identical in all their dimensions. The mode of 

 viewing the needle is of great consequence. The eye must be placed in such 

 a situation that the four objects concerned shall be seen in the same vertical 

 line ; namely, the degree on the lower circle, the point of the lower bar, the 

 corresponding degree on the upper circle, and the point of the upper bar. The 



