MAGNETIC SHIELDING IN HOLLOW IRON CYLINDERS. 651 



stages of induction, determined by the direction of the magnetic force used to de- 

 magnetise the iron by decreasing reversals. 



§ 32. Shields A and B were now reverted to, and in order to wipe out as far as 

 possible the previous magnetic history of the iron, the shields were demagnetised by 

 repeated reversals of both the circular and the longitudinal magnetising forces. The first 

 four columns of Table XIII. give the measurements obtained due to increments of the 

 circular magnetising force, following upon demagnetisation leaving the iron without 

 apparent polarity, under the influence of decreasing reversals of the longitudinal field 

 last applied. The second four columns, on the other hand, give the measurements due 

 to increments of the circular magnetising force following upon demagnetisation obtained 

 under the influence of decreasing reversals of the circular field last applied. In the first 

 case the iron had been demagnetised by a force at right angles to the subsequent 

 magnetising force, and the results are plotted as dotted line curves in fig. xviii. for both 

 qualities of iron. In the second case the iron had been demagnetised by a force acting 

 in the same direction (positive and negative) as the subsequent magnetising force, and 

 the results are plotted as full line curves in the same fig. for both qualities of iron — 

 these curves being the normal B-H carves of induction. The differences between the 

 two curves are sufficiently well marked, confirming the results obtained under the con- 

 ditions described in the section immediately preceding, for a different quality of iron. 



§ 33. We may therefore conclude that during the early stages of induction iron is 

 more permeable to a re-application of a magnetising force in the same direction (whether 

 positive or negative) as that used in the immediately preceding demagnetising process, 

 than it is to a force (whether positive or negative) at right angles to that used in the 

 immediately preceding demagnetising process. The maximum difference did not exceed 

 500 C.G.S. magnetic units, but as this occurs at low values of the B-H curve, it 

 amounts to a very considerable proportion of the induction. The full line curves 

 exceed the dotted curves by about 30 to 35 per cent., and this maximum corresponds 

 approximately with the period of maximum permeability (cZBcZ/H) in the specimens of 

 iron used. When, however, the magnetising force is increased, the two curves rapidly 

 approximate and appear to become one. 



Theoretical Considerations. 



§ 34. Ewing's modification of Weber's theory of magnetic induction assumes (with 

 Weber) that the molecules are magnets pointing in all directions in unmagnetised iron, but 

 falling into line and showing polarity in the mass when a magnetising force is applied. 

 It dispenses with any arbitrary forces controlling the molecular magnets, and explains 

 that the constraint which " proceeds only from their mutual action as magnets evidently 

 suffices to explain generally the characteristics of the magnetising process." In order 

 to arrive at some possible explanation of this directional permeability at the early 

 stages of induction, consider the demagnetisation of iron by decreasing reversals. If 



