200 



LORD ItAYLEIGIT. 



So long as the iron can be treated as idéal, the above formula holcls 

 good, and upon the supposition of a sufficiently sudden break there 

 seems to be no reason why it sliould not afford a tolerable approxima- 

 tion to the actual maximum value of V. The proportionality between 

 spark-length and primary current was found to hold good in Walter/s 

 experiments over a considérable range. 



When the core is very long in proportion to its diameter, or when 

 it approximates to a closed circuit, the behaviour of the iron may 

 deviate widely frorn that described as idéal, and the quantity denoted 

 by L lias no existence. But the princijjle remains that the energy of 

 charge at the moment preceding the secondary spark cannot exceed, 

 though it may somewhat closely approach, the energy of the primary 

 current before break. 



We hâve next to consider how the energy of the primary current is 

 to be reckoned, and here we encounter questions as to which opinion 

 is not yet undivided. The gênerai opinion would, I suppose, be that the 

 bodily magnetization of the iron represents a large store of available 

 energy. If this be correct, the inference would be irrésistible in favour 

 of a very long, or a completely closed, iron core. Some years ago 

 reasoning on the basis of the theory of Wauburg and Hopkinson, I 

 endeavoured to show that highly magnetized iron could not be regarded 

 as a store of energy — that the energy expended in producing the mag- 

 netization was recoverable but to a s mail extent, or not at ail. Although 

 this conclusion does not appear to have been accepted, perhaps in 

 conséquence of an erroneous application to alternating current trans- 

 formers, 1 still see no means of escape from it. The available energy 

 of a highly magnetized closed circuit of iron is insignificant. If the 

 length be limited, there is available energy, in virtue of the free pola- 

 rity at the ends. 



The theory is best illustrated by the case of an ellipsoid of révolution 

 exposed to uniform externat magnetizing force Jjî>' acting parallel to 

 the axis. „If 3 De l ne magnetization parallel to the axis of symmetry 

 {2 c), the demagnetizing effect of 3 i s N 3> where N is a numerical 

 constant, a function of the eccentricity (e). When the ellipsoid is of 

 mie ovary or elongated forin, 



J ) „On the Energy of Magnetized Iron", Phil. Mag. 22, p. 175, 1886; 

 Scientific Paper s, II, p. 543. 



