Jan. 23, 1890] 



NATURE 



273 



MAGNETISM} 



II. 



"\^HEN one considers that the magnetic property is 

 ^^ peculiar to three substances — that it is easily- 

 destroyed by the admixture of some foreign body, as 

 manganese — one would naturally expect that its existence 

 would depend also on the temperature of the body. This 

 is found to be the case. It has long been known that iron 

 remains magnetic to a red heat, and that then it somewhat 

 suddenly ceases to be magnetic, and remains at a higher 

 temperature non-magnetic. It has long been known that 

 the same thing happens with cobalt, the temperature of 

 change, however, being higher ; and with nickel, the tem- 

 perature being lower. The magnetic characteristics of 

 iron at a high temperature are interesting. Let us return 

 to our ring, and let us suppose that the coils are insulated 

 with a refractory material, such as 

 asbestos paper, and that the ring is 

 made of the best soft iron. We are 

 now in a position to heat the ring to a 

 high temperature, and to experiment 

 upon it at high temperatures in exactly 

 the same way as before. The tempe- 

 rature can be approximately deter- 

 mined by the resistance of one of the 

 copper coils. Suppose, first, that the 

 current in the primary circuit which 

 we use for magnetizing the ring is 

 small ; that from time to time, as the 

 ring is heated and the temperature 

 rises, an experiment is made by re- 

 versing the current in the primary cir- 

 cuit, and observing the deflection of 

 the galvanometer needle. At the or- 

 dinary temperature of the air the de- 

 flection is comparatively small ; as the 

 temperature increases the deflection 

 also increases, but slowly at first ; when 

 the temperature, however, reaches 

 something like 600° C, the galvano- 

 meter deflection begins very rapidly 

 to increase, until, with a temperature 

 of 770^ C, it attains a value of no less 

 than 1 1,000 times as great as the de- 

 flection would be if the ring had been 

 made of glass or copper, and the same 

 exciting current had been used. Of 

 course, a direct comparison of 11,000 

 to I cannot be made : to make it, we 

 must introduce resistance into the 

 secondary circuit when the iron is 

 used ; and we must, in fact, make use 

 of larger currents when copper is 

 used. However, the ratio of the induc- 

 tion in the ca-e of iron to that in the case of copper, at 

 770° C, for small forces is no less than 11,000 to I. Now 

 mark what happens. The temperature rises another 

 15^ C. : the deflection of the needle suddenly drops to a 

 value which we must regard as infinitesimal in comparison 

 to that which it had at a temperature of 770° C. ; in fact, 

 at the higher temperature of 785° C. the deflection of the 

 galvanometer with iron is to that with copper in a ratio 

 not exceeding that of ri4 to 1. Here, then, we have a 

 most remarkable fact : at a temperature of 776° C. the 

 magnetization of iron 11,000 times as great as that of a 

 non-magnetic substance ; at a temperature of 785'^ C. 

 iron practically non-magnetic. These changes are shown 

 in Y'x'g. 8. Suppose now that the current in the primary 

 circuit which serves to magnetize the iron had been great 

 instead ot very small. In this case we find a very differ- 



' Inaugural Address delivered before the Institution of Electrical En- 

 gineers, on rhursday, January 9. by J. Hopkinson, M.A., -D-Sc, F.R.S., 

 President. Continued from p. 254. 



ent order of phenomena. As the temperature rises, the 

 deflection on the galvanometer diminishes very slowly 

 till a high temperature is attained ; then the rate of 

 decrease is accelerated until, as the temperature at 

 which the sudden change occurred for small forces 

 is reached, the rate of diminution becomes very 

 rapid indeed, until, finally, the magnetism of the 

 iron disappears at the same time as for small forces. 

 Instead of following the magnetization with constant 

 forces for varying temperatures, we may trace the curve 

 of magnetization for varying forces with any temperature 

 we please. Such curves are given in Diagrams 9 and 10. 

 In the one diagram, for the purpose of bringing out 

 different points in the curve, the scale of abscissae is 20 

 times as great as in the other. You will observe that the 

 effect of rise of temperature is to diminish the maximum 

 magnetization of which the body is capable, slowly at 



Wrought Iron 

 MAGNETisiNa Force 0-3. 



100 200 300 400 SOD 600 700 78S800'C 



Fig. 8. 



first, and rapidly at the end. It is also very greatly to 

 diminish the coercive force, and to increase the facility 

 with which the body is magnetized. To give an idea of 

 the magnetizing forces in question, the force for Fig. 8 

 was 03 ; and as you see from Figs. 9 and 10, the force 

 ranges as high as 60. Now the earth's force in these 

 latitudes is 043, and the horizontal component of the 

 earth's force is o'i8. In the field of a dynamo machine 

 the force is often more than 7000. In addition to the 

 general characteristics of the curve of magnetization, a 

 very interesting, and, as I take it, a very important, fact 

 comes out. I have already stated that if the ring be sub- 

 mitted to a great current in one direction, which current 

 is afterwards gradually reduced to zero, the ring is not in 

 its non-magnetic condition, but that it is, in fact, strongly 

 magnetized. Suppose now we heat the ring, whilst under 

 the influence of a strong magnetizing current, beyond the 

 critical temperature at which it ceases to have any mag- 

 netic properties, and that then we reduce the current to 



