53G 
MR. J. S. TOWNSEND ON MAGNETIZATION OF LIQUIDS. 
cause a current to leak to the secondary if good non-conductors are not used to 
separate the two circuits. 
It is necessary that A, B, and C should be at long distances apart, so as to satisfy 
the two following conditions:— 
1 . The primary circuit of B must have no lines of force through the secondary 
of A. 
2 . Changing the number of coils in the primary of 0 should have no effect on the 
galvanometer when C is cut out of the secondary circuit (by joining x and y ). 
Details of Apparatus. 
The primary circuit of the solenoid contained 1509 turns, wound in four layers on 
a cylindrical glass jar 47 centimes, long and about 4 centims. internal radius. The 
secondary, of 841 turns, was carefully insulated from it b} r means of several layers of 
paraffin paper, and occupied a length 4 centims. each side of the centre. 
The mean internal sectional area at the centre perpendicular to the axis of the 
tube was found by observing with a cathetometer the rise in level of water in the 
tube, when a known additional volume was added to it. 
The area thus obtained was 50’6 sep centims. 
Let a = radius and l half the length of the solenoid AB. The force F due to unit 
current is less than 4irn by the amount contributed by the two productions AA' and 
BB' of the solenoid, A' and B' being theoretically at infinity. 
A' j — n —P F + n 
The solenoids, A'A and B'B, can be replaced by uniform surface distributions of 
magnetism, — n and + n, over the plane circular ends at A and B, as far as forces 
at places external to A'A' and BB' are concerned, n being the number of turns per 
centirn. # 
Hence the force F at the centre of the solenoid is less than 4mi by that due to a 
pole, — 7 ma 2 at A and + nna 2 at B, so that 
F=4 TO (l-A). 
The second correction for the ends of the solenoid is found by considering the 
induction through the secondary circuit due to the magnetism induced on the surface 
of the liquid. In cases where k is small this correction is easily made since terms 
involving P may be neglected. 
* Maxwell, ‘ Electricity and Magnetism,’ vol. 2, sect. 676. 
