234 HENRY A. EOWLAND 







same. In such a case as that we leave the magnet about as strong as 

 it was before. But that would not be the case if we compress magnet- 

 ism until we get it up to the point of magnetization of the centre. In 

 that case we should need the whole mass, and it is almost impossible 

 to magnetize to any extent without the centre coming in. It depends 

 on the length of the bar. If we bring the bar around, making a com- 

 plete magnetic circuit of the thing, so that the lines of force do not 

 have to pass out into the air at all when we put a wire around it so as 

 to wind it like a ring at every point, in that case the whole cross-section 

 becomes equally magnetized, if it is not bent too much. If it is a large 

 ring of small cross-section, it is perfectly magnetized across from side 

 to side. We know that perfectly well; it is a result of the law of con- 

 servation of energy. The case of dynamos is like that. We require 

 the whole cross-section to transmit these lines around. The resistance 

 to the magnetization comes partly from this opening and partly from 

 the iron. We have no gain in making these cylinders hollow; indeed 

 we rather increase the outside surface to let lines of force flow into the 

 air. In the case of a dynamo machine, the solid form is not only 

 desirable, but by far the most efficient. 



I have thought of that matter a great deal, and experimented upon 

 it. Indeed this closed circuit is the very idea from which the permea- 

 bility of the iron is determined. All the calculations upon that sub- 

 ject are based upon that law. I think there can be no doubt that in 

 the dynamo the solid form is the proper form, and that the whole cross- 

 section is effective. The whole cross-section of a round piece is just as 

 effective as the whole cross-section of a flat piece. The flat piece ex- 

 poses more surface to the air, and there is more surface for the force 

 to escape from. That is another reason for not making the magnets 

 flat. The round form is that in which there is the least surface, and 

 therefore the least liability of the lines of force to escape. You can 

 conduct the lines of force by a round piece to any point you desire much 

 better than by a flat piece. 



[Discussion by Professor Sylvanus Thompson.] 



I do not know that the theory bears upon the solidity of the core. 

 Of course, the more iron in there the better is the efficiency of the 

 machine. I suppose there would be no objection to dividing that 

 cylinder up into a number, so that the Foucault currents could not 

 exist, if the exterior form was round; but I do have an objection to 



