538 
MR. J. S. TOWNSEND ON MAGNETIZATION OF LIQUIDS. 
When a current flows through a solenoid, and magnetizes a cylindrical body of 
equal length whose value for k is small, there are uniform distributions +, and —, 
2 7rnk induced on the plane ends, and of the total quantities fl-, and —, 47 rnk (A) 
induced, one-half resides on the ends and the other half on the cylindrical surface. 
This latter distribution could not be got rid of by producing the ends of the 
cylinder heyond the magnetizing coil. 
The actual value of the normal force N at a point distance y from the end of a 
solenoid of circular section of radius a is given by the equation 
N = 
2f + 4a. 2 _ 2 vb/ 2 + 4« 2 E 
_a \/ ?/ 2 + 4« 2 a 
V, 
K and E being the complete elliptic integrals of the first and second kind, the 
modulus k being' 
2 a 
s/tf + 4« 2 
(N may also be found by differentiating the expression for Y in Thomson and Tait’s 
‘ Natural Philosophy,’ Ex. If., Section 546). 
Making all the corrections, it is found that the change in induction m is 
O ’ O 
47 rk 47rnN / . A 
1 
a 2 _B /-0530V 
2l 2 ~ A \ 77 J 
• Substituting the following values, 
n 
we get 
1 5 0 9 . 
4 7 5 
therefore 
N' = 841; A = 50-6 ; a = 4*5 ; 2^ = 47; B = 93 ; 
m = k X 2155-10 5 (l — -049), 
7 m 
’ = 205T0 6 ’ 
The adjustable inductance was made in two separate parts, c 1 and c 2 , shown in 
fig. 2. The larger changes in induction were made by c v the secondary circuit of 
which consisted of seven turns wound round the inner bobbin, 18' 61 centims. in 
diameter, and had its terminals at t v 
On the outer surface of the wooden ring R, 37’17 centims. in diameter, were wound 
10 turns of wire of small diameter close beside one another. The ends of 1, 3, and 6 
of these circles of wire were twisted together and terminated in 6 mercury cups. 
The rirg R and the bobbin were placed in concentric and coplanar positions, and were 
then fixed rigidly together by means of ebonite bars. 
