126 Rev. Mr Scoresby on the Uniform Permeahility of 



cuity and practice, to consider this application of the principle 

 under three separate cases. 



Case I. — Where the level oj Ike mines is known to be the same, but 

 the direction is doubtful. 



In this case it is only necessary to apply the magnets in one 

 of the mines in a given position, and then at the same horizontal 

 level of the other, (say on the floor of the mine, or at any par- 

 ticular distance above), to try the deviations of the compass at 

 moderate intervals, from side to side, directly across the head- 

 way, when, from the observation of the greatest deviation, the 

 point nearest to the magnets may be ascertained. One precau- 

 tion, however, will be necessary. If the direction of the com- 

 pass-needle be not exactly at right angles to the position of the 

 magnet, then the greatest deviation will not occur at the nearest 

 point but a little beyond, towards the place where the action of 

 the magnet is most perfectly tangential *. Hence, it will be 

 generally advisable, not only to adjust the needle of the compass, 

 by a directing magnet, as near as may be to a right-angular 

 position with the magnet (as shown by the known magnetic 

 position of each mine), but also to repeat each experiment with 

 both poles of the magnet alternately presented to the compass, 

 so as to obtain the difference of deviations of the two poles 

 at each position of the compass. In order to conduct this ex- 

 periment with the greater facility and uniformity of effect, the 

 simple apparatus represented in Fig. 7, Plate V. of preceding 

 volume, formed out of a piece of board, may be useful. 



C is the compass, and D a directing magnet sliding in a groove, 

 which is so adjusted as to its distance from the compass, that 

 the needle may assume a position parallel to the straight edge 



• As this circumstance might in some cases occasion a considerable error, 

 if only the deviation by one pole were taken, it would be useful to have the 

 means of ascertaining the extent of error capable of being thus produced. 

 This would be given by a table of the powers of the magnets, calculated from 

 the ^nes of deviation instead of the tangents^ which would shew the distance 

 of the magnet from the compass, as placed at right angles to the needle in 

 its deviated position. At great distances, however, no material mistake can 

 arise from the neglect of this source of error, — the powers of the sines and 

 tangents being so nearly "the same in deviations under 5 or 6 degrees as to be 

 within the ordinary limits of error of observation. 



