128 AKSEL S. STEEN. TERRESTRIAL MAGNETISM. [NORW. POL. EXP. 
side of the needle outwards, and as naturally no inversion of the poles could 
be performed, since the needle was also to be employed for determinations 
of the intensity, the mean value of the inclination found by the method 
described above is only relative, and must be corrected for the index-error 
of the needle in question. 
DETERMINATION OF THE INDEX-ERROR. 
If the inclination observed with the apparatus is called J’, the actual 
inclination J, and the index-error of the needle used 4, we have 
CT A, 
This index-error is due to two different sources, namely, first, the cir- 
cumstance that the magnetic axis of the needle does not coincide with its 
geometrical axis, and secondly, that the needle’s axis of rotation does not 
pass exactly through its centre of gravity. 
Fig. 3. 
In fig. 3, NS indicates the needle’s geometrical axis, N’S’ its magnetic 
axis, the angle between them being called ¢. @ is the needle’s centre of 
gravity, and O the point of intersection of the axis of rotation with the 
vertical plane. The line connecting these two points, 7, then makes with 
