Magnetic Permeability of an Iron Bar. 285 



cut in the lathe without being accurately shaped into true 

 planes, and that this film should continue to be present when 

 the pieces are pressed together by external load. It is less 

 easy to see how so considerable a film of air should still be 

 found when the surfaces are true planes, before they are 

 pressed together. This, however, is perhaps possible enough ; 

 and the idea that the resistance is due to an actual film of air 

 gains much probability from the fact that when truly plane 

 surfaces are forcibly pressed together the cut bar behaves 

 almost exactly as if it had not been cut. On the other hand, 

 it is difficult to reconcile the idea that the " resistance" of a 

 joint is due to an air film, with the fact which these experi- 

 ments clearly demonstrate — that the " resistance" diminishes 

 greatly as a state of magnetic saturation is approached. We 

 have seen that this diminution takes place under conditions 

 which make it apparently impossible that the diminution can 

 have been due to any real increase of closeness in the contact 

 of the air-surfaces. In the case of a rough-cut bar strongly 

 compressed we cannot suppose that the actual thickness of 

 the air-space suffers any material reduction as the bar becomes 

 magnetized. If we assume the resistance of the joint to be 

 due to the simple presence of this air-space, the alternative 

 would be that the permeability of air, or at least of air in 

 the condition in which it exists when condensed on the sur- 

 faces of bodies, increases under the action of strong magnetic 

 forces to a very notable extent, a conclusion too startling to 

 be accepted on the evidence of these experiments alone. 



"Whatever be the explanation of the results, they make it 

 clear that a joint between two portions of the iron core of a 

 magnet has in general a distinct magnetic "resistance," 

 which is in all cases a function of the magnetic induction, 

 decreasing much when the induction is increased. That this 

 resistance is present even when the surfaces of the joint are 

 carefully faced to be true planes ; being in fact about as great 

 then as when the surfaces are rough, if the magnetizing force 

 is small, though less than when the surfaces are rough, if the 

 magnetizing force is great. That when the surfaces are rough 

 the resistance is somewhat reduced, but by no means wholly 

 eliminated, by using force to press the parts together. That 

 when the surfaces are true planes a considerable force will 

 destroy the resistance of the joint almost completely. And, 

 finally, that this destruction of the resistance by pressure, 

 when the surfaces are true planes, is only a very little less 

 complete when a film of gold-leaf is interposed between the 

 iron faces. 



