1895.] on the Effects of Electric Currents in Iron } &c. 551 



confusing. If we had done so, we should have seen that the 

 diminution of disturbance with increase of speed was less with 

 No. 2 coil than with No. 1 coil, and that it had disappeared entirely 

 with No. 3 coil, and we should further see that the current in No. 2 

 coil lay behind No. 3, and in No. 1 behind No. 2. I can show you 

 the contrast more effectively with the Westinghouse transformer 

 with a divided core. This is now connected to the lowest gal- 

 vanometer. We turn slowly : notice the deflection. We turn faster : 

 you see the deflection is increased instead of diminished, as it was 

 with the central coil of the solid electromagnet. Experiments such 

 as these are at once applicable to transformer cores and the cores of 

 dynamo machines. They show that in practice manufacturers have 

 divided the iron about enough, and not too much. 



Fig. 14.— Magnetic Curve-Tracer Curves for Soft Iron Bars. 



(a) Cycle performed slowly. 



(b) Period of cycle 3 seconds. 



(c) Period of cycle 0*43 second. 



I have here on the table an instrument designed by Prof. Ewing 

 for the purpose of describing the curves which express the relation 

 between induction and magnetising force in iron. It served in 

 Prof. E wing's hands to illustrate the point which we have been 

 discussing. I will throw upon the screen curves taken from Prof. 

 Ewing's paper read before the Royal Society. The curves have 

 been taken from solid samples of iron — I mean iron which is con- 

 tinuous, and not divided for the purpose of annulling the current 

 in it. In the first curve (a, Fig. 14) we have the result in which 

 the cycle has been passed through very slowly, and is the true curve 

 of magnetisation. We have b, where the cycle has passed through 

 in three seconds — and you will observe that the amplitude of the 



2 p 2 



