Mr. G. Hookham on Permanent Magnet Circuits. 187 



ends carefully refitted by filing and scraping. If, now, this 

 ring is once strongly magnetized by an electrical current, we 

 obtain a powerful permanent magnet. The magnetic circuit, 

 however, is anything but complete. There are two breaks in 

 it ; very minute, no doubt, if their mere dimensions are taken 

 into account ; but recent experiments seem to show that, how- 

 ever carefully we may fit together surfaces of iron, there is 

 still an easily measurable resistance which is greater than 

 would be due to the dimensions of the air-space ; that there 

 is in fact, in addition, a surface-resistance. Such a resistance 

 is the equivalent of a small current tending to reverse the 

 magnetism of the iron. Now, what enables soft and pure 

 iron to withstand this demagnetizing agency? Apparently 

 nothing but the coercive force of the iron, small as it is. 

 Once magnetized, the softest iron under any circumstances 

 remains at ordinary temperatures a permanent magnet, how- 

 ever weak ; and one can only suppose that it is in virtue of 

 this very small coercive force that the soft iron magnet over- 

 comes even the small resistance of well-fitted joints. If this 

 supposition were well founded, I felt almost certain that when 

 very hard magnets of tungsten steel were substituted for the 

 soft iron, air-space resistances incomparably greater might be 

 overcome ; and yet the magnetic intensity in the circuit be 

 very considerable, and to all intents and purposes absolutely 

 constant. It was at any rate quite certain that this state of 

 things would obtain, if the power of overcoming resistances in 

 the circuit were at all proportional to the coercive force of the 

 material of the magnets. My anticipations have been com- 

 pletely realized, and I have been able to construct magnets 

 which are truly permanent. 



As to the special arrangement, when the object in view, as 

 in this case, is subject to commercial considerations, cost of 

 production makes the use of straight bar-magnets almost 

 imperative. A number of these are enclosed in a brass tube, 

 to the ends of which are attached soft cast-iron arms, curving 

 round towards the pole-pieces, which closely face each other, 

 so as to form a narrow sht in which the thin disk-armature of 

 the meter revolves. The magnetic resistance of the arms is 

 practically negligible. As to proportions I was, in principle, 

 guided by the analogy of the voltaic circuit. There, if we 

 wish to overcome a considerable external resistance, the plan 

 is to pile up the electromotive force by increasing the number 

 of cells in series, paying little regard to the consequent in- 

 crease of internal resistance. Here the steel masnet-bars 

 constitute the magnetic cell, whose magnetomotive force — to 

 use what is at any rate a most convenient expression — is 



