Component of the Eartlis Magnetism at Oxford. 445 



not likely to be troublesome ; but still it is as well to have 

 some idea of its amount. 



In determining this I have used the direct method. I have 

 wound 100 turns of wire about the middle of the magnet, and 

 joined this up with the circuit of the ballistic galvanometer, 

 In the same circuit, and so arranged that the magnet can be 

 inserted, are the four coils used to determine the number of 

 lines of force of the magnet. The vertical component of the 

 earth's magnetism is used, as being larger and more con- 

 venient than the horizontal. The magnet standing upright 

 is first turned upside down. This produces a deflection due 

 to 200 times the induction. The magnet is then reversed, 

 and half the difference of the readings is taken. The magnet 

 is then inserted in the four coils, which gives a deflection due 

 to four times the number of lines of force in the magnet. It 

 is then taken out again, and half the difference of the readings 

 taken. 



In my determination the deflections were as follows : — 



OAnv . , ,. 4 X total lines ratio induction by H 



200 X induction. «/. „ KKZTl — j^~ — 1 • 



ol iorce. oU X tan dip magnet power 



I. . . 96'-0 306'-5 '00257 



II. . . 117'-0 339'-0 -00283 



The observations were very difficult, especially with mag- 

 net I. It seemed as if at a certain moment spontaneous 

 alterations began to take place in the magnet, which kept the 

 scale of the galvanometer flying about so that no observations 

 could be made. I thought this might be due to variations of 

 the terrestrial force, and laid the magnet at right angles to 

 the magnetic meridian ; but the changes went on as before 

 though their direction was altered. It appears therefore that 

 when such a magnet is exposed to such a force as that of 

 the earth's magnetism, the charge continues for some time 

 to work its way irregularly among the non-homogeneous 

 material of the magnet, and on withdrawal of the force its 

 effect subsides in the same way. 



The above expression for the effect of terrestrial induction 

 is most convenient for the purpose of such calculations as 

 those we are dealing with. But it is of more general interest 

 to express this effect by means of the quantity magnetic 

 resistance, which I have introduced to denote the quotient of 

 magnetic potential by lines of force per unit surface, or mag- 

 netic induction. These quantities are, as deduced from the 

 above determinations, 



Magnet. Resistance. 



I -743 centim. 



II -685 „ 



Phil. Mag. S. 5. Vol. 17. No. 108. June 1884. 2 H 



