292 



Messrs. T. Gray and A. Gray. 



[Feb. 14, 



regulate tlie sensibility, as then one of the needles must be magnetized 

 and the other demagnetized to a greater or less extent, depending on 

 the position of the magnet. According to this latter arrangement, 

 if we suppose the needles to be parallel or nearly so, and I to be 

 the magnetic field intensity at the upper needle, I' that at the lower 

 needle in the same direction, m the magnetic moment of the upper 

 needle, m' that of the lower needle, C the current flowing, 6 the 

 deflection produced, and K a constant, we have — 



n -f-rlm — I'm 7 , n 



C=K tan 6. 



m-\-m 



Slight changes in m or w! or in both may, therefore, affect the con- 

 stant of the ordinary instrument very seriously, and as a matter of 

 fact its constant has to be continually redetermined, and instead of 

 preserving a practically constant zero, it is very sensitive to magnetic 

 disturbances in the neighbourhood. 



In the case, however, of needles adjusted to be accurately vertical 

 these disadvantages do not exist. The needles cannot be affected in 

 the same way by directing magnets and retain their astaticism for 

 uniform field, I, I' being the horizontal field intensities at the upper 

 and lower extremities of the needles, the current strength, 6 the 

 deflection of the needles, and K a constant depending on the coils, 

 we have approximately — 



C = K(I-r) sin0. 



The sensibility of the instrument can, therefore, be increased to 

 any desired extent by placing the magnet M at a greater distance 

 from the needles (or by counteracting its action by a smaller magnet 

 placed nearer to the needles) so as to make I — V sufficiently small. 

 Further, variations of the strength of the horseshoe needles produce 

 no effect unless they consist of changes of magnetic distribution, 

 which may produce a deviation from perfect astaticism. When the 

 instrument is properly adjusted and the needles are as nearly as 

 possible uniformly magnetized, but little disturbance of this kind can 

 be produced by the magnetizing action of the coils, since both poles of 

 each have their magnetism augmented or both diminished at the same 

 time in the arrangement of fig. 2, or both poles of one are magnetized 

 more intensely in some degree, and both poles of the other weakened 

 in the arrangement of fig. 1. 



Another possible arrangement of such a system of needles is with 

 like poles turned in similar directions. The system will still be per- 

 fectly astatic if properly adjusted ; and to give a return couple 

 towards a zero position a magnet may be used, placed, for example, 

 horizontally in the vertical plane at right angles to the front of the 

 instrument, in a line passing through the suspension thread. If this 



