662 MR JAMES RUSSELL ON 



superposition under which theoretical conditions are for all practical purposes fully 

 satisfied. 



The diagrams given under fig. xxxiv. are intended to represent graphically first the 

 increase and then the decrease of the shielding ratio under the TC conditions, (l) shows 

 the iron shield under the influence of the transverse field only ; (2) the superposition of 

 a sufficient circular force to produce zero magnetisation at n, with an increase of the 

 shielding ratio ; (3) a further increase of H c and a further increase of the shielding 

 ratio. In (4), however, a further increase of H c lowers the shielding ratio. It is an 

 interesting deduction from these lines of force diagrams that, when the flux at n has 

 changed sign and is not less than the total flux across the shielded space, the 

 magnetic lines of the transverse field cannot be represented as being continuous 

 with the magnetic lines crossing the shielded space. See (3) and (4), also fig. xxxv. 

 (2). These lines of force diagrams are limited to a representation of the changes which 

 occur in the shell of, and the internal shielded space within, the shields. Owing to the 

 few number of lines that can be used without a sacrifice of clearness, they do not show 

 the variation of the number of magnetic lines converging upon the iron which we 

 have seen follows upon the superposition of the circular magnetising force. Du Bois' 

 lines of force diagrams (loc. cit.) may be referred to as showing the effect which various 

 thicknesses of the cylindrical shells have upon the number of magnetic lines converging 

 upon the iron. 



We therefore conclude that when increments of the circular field are superposed 

 upon a pre-existing induction due to the transverse field (TC conditions) the shielding 

 ratio approximates nearer and nearer to 



, i d dB. 



as the value of the transverse field is reduced. Thus the shielding ratio minus unity 

 under the above conditions of field superposition is within the limits of experimental 

 error proportional to the permeability impressed upon the iron by the superposed 

 circular field ; n being defined as the differential permeability dBJdii,. and not the ratio 

 permeability as was found to be the case when the transverse field alone is acting upon 

 the iron (§ 11). 



Conclusions under CTT conditions. 



§ 49. Reference to figs. xx. and xxi. shows that the shielding ratio curves under 

 the CTT conditions differ from those obtained under the TC conditions, and that when 

 various values of the transverse field are used, figs. xxvi. and xxvu., the differences 

 which exist in the initial values of the shielding ratios diminish at a much slower rate as 

 H c increases under the CTT as compared with the TC conditions. No apparent connec- 

 tion appears to exist between the shielding curves, which descend from their maximum 

 values in the vertical axis as the circular magnetising force increases, and the values of 

 dB/dll, which increase as the circular magnetisation creeps up during the repeated 



