Terrestrial Magnetism. 169 
The mean values of 8 and @, as deduced from these four observations, are 
=//02. 00) 7 ed — Ure o.6 3 
and if we take the force corresponding to these mean values as unity, the numerical 
values of the force, derived from each separate observation, will be obtained by sub- 
stituting the corresponding values of 6 and 6, given by the preceding table, in the 
formula 
log ¢=log cos 6—log cos 6 + log sin (8, —0,) —log sin (¢—6). 
We find in this manner 
$1=-9994, $2=1.0000, g:=1.0011, ¢,=.9995. 
The greatest deviation of these values from the mean is .0011; so that the limit of 
error in the determination of the force, would appear from these observations to be 
little more than the one-thousandth part of the entire quantity. 
It appears then from the preceding, that the Terrestrial Magnetic Intensity may be 
determined, together with the Dip, with the aid of a single instrument, and by a pro- 
cess even somewhat less troublesome than that by which the dip alone is usually de- 
termined.* To any one who considers the numerous precautions required, in the 
common method of determining the magnetic intensity, the saving of time and labour 
thus effected will be abundantly obvious. But it is an advantage of much greater 
moment, that the results of the proposed method, so far as the intensity is con- 
cerned, will be less liable to error than those obtained in the usual manner, as long as 
the dip exceeds 45°; and that, in our latitudes, the accuracy of the new method is 
nearly three-fold that of the old. 
* The same number of readings is taken in the two cases, while in the proposed method the process of 
reversing the poles is dispensed with. <— 
—-~ 
