444 Prof E. Wilson. The Distribution of Magnetism. 



as great as in fig. 6. Apart from phase-displacements we see that the 

 interior portions of the cylinder would be quite useless with regard 

 to induced magnetism, unless very large or very small external 

 magnetising forces were employed. 



VII. Experiments were made upon a steel wire O01 inch (0'00254 

 cms.) diameter, the object being to discover if a true time lag in 

 magnetism exists.* Frequencies of 5, 72, and 125 periods per second 

 were tried, and the results obtained were compared with the curve of 

 magnetic hysteresis obtained by the Ballistic-galvanometer method. 

 The curves so obtained would indicate a somewhat higher dissipation 

 of energy with the high frequencies. Now the effects shown in fig. 3 

 would also be observed in a wire O01 inch diameter if the frequency 

 were 600. The inference is that at 125 periods per second there would 

 still be disturbances due to induced currents. The author inclines to 

 the opinion that the effects associated with time lag in magnetism 

 may be due to induced currents in the wire. 



VIII. Lord Kelvin has computed that the earth's magnetism is 

 travelling in the direction of the sun round the earth with a periodic 

 time relatively to the earth of 960 years. The magnetic declination 

 at London had an amplitude of 24 34' 30" W. in 1820. Diurnal 

 variation is observed to the extent of some minutes of arc. The 

 reaction of the induced currents in the cylinder produces a displace- 

 ment of polarity. In a cylinder similar in all respects to the one 

 experimented upon, but having a diameter equal to that of the earth, 

 a periodic time of 960 years would produce similar magnetic and 

 electric events as would be observed in the 10-inch cylinder if 

 the latter could be rotated with a periodic time 12 x 10~ 6 second. 

 This is nearly 2,000,000 times as fast as the fastest speed in these 

 experiments. Diurnal changes in spite of small magnetic force would 

 produce effects confined to the surface of such a cylinder, and would 

 not appreciably disturb an effect due to a periodic time of 960 years. 

 Table II shows that when the intensity of magnetic induction at the 

 surface of the 10-inch cylinder has the value 169, phase differences of 

 the order 44" (360 == 1 period) are experienced with a periodic time 

 of 45 seconds. The magnetic force H per centimetre linear in C.G.S. 

 units would be about 0'5 in this iron, giving an average permeability 

 of about 280. This is a force of the same order as the earth's total 

 intensity. 



IX. If we were to work upon smaller and smaller masses of iron 

 until we reached the limit at which all the properties of the original 

 mass were still preserved, we should require enormous speeds of 

 rotation to produce the disturbances examined in this paper. Suppose 

 the diameter of such a mass were of the order O g OOO,0001 cm., then 

 to produce similar effects to those which we have observed at 90 seconds 



* See Hopkinson, Wilson, and Lydall, ' Roy. Soc. Proc.,' TO!. 53, p 352. 



