applied to Terrestrial Magnetism. 459 
If the preceding views be correct, it will follow that, where the dip is considerable, 
we cannot hope to determine with any accuracy the relative values of the total in- 
tensity, by the observation of its horizontal component, unless the dip be observed at 
the same time at each place. Still, however, the final values of the intensity will be 
affected, to a large amount, by the errors of observation to which the results obtained 
with the dipping needle are liable. The chief of these errors is that due to the fric- 
tion of the axle on its supports ; and it has been already shown* that if the amount of 
that error in the natural position of the needle be denoted (in parts of radius) by e, 
the induced error in the determination of the force by the receiyed method, will be 
e tan 0; 
while the error, arising from the same cause, in the needle, is 
e 
; sin (6—@) ~ 
Accordingly, when 6—6=90°, or when the position of the needle when loaded is 
perpendicular to the line of the dip, the error in the determination of the force is 
reduced to e ; and is less than the corresponding error in the common method in the 
ratio of unity to the tangent of the dip. 
* p. 167—8. 
VOL. XVII. 41 
