6G6 MR JAMES RUSSELL ON 



I. (c) Shielding Ratios with Transverse and Longitudinal Fields. 



§ 53. Table XX. supplies measurements for both shields under the four conditions of 

 field superpositions, the value of the transverse field being H t = 20'9 C.G.S. units. The 

 letters TL, TLL, LT, and LTT have the same signification as in § 38, L (longitudinal) 

 taking the place of C (circular). The sequences observed in these measurements are also 

 those which were observed in the measurements recorded in Tables XV. and XVI. 

 (figs. xx. and xxi.) and detailed in Table XIV. By substituting L and E { for C and H c 

 respectively, they are at once made fully applicable for the determinations given in 

 Table XX. The curves are plotted as full lines in fig. xxxviil for shield A, and in 

 fig. xxxix. for shield B. They differ from fig. xx. and xxi. only in the fact that the 

 longitudinal has replaced the circular magnetising force. 



§ 54. It was observed that the origin of the new curves in the vertical axis was 

 lower for both shields by at least unity, when compared with those previously discussed 

 (figs. xx. and xxi.). The only known difference between the magnetic conditions of the 

 shields appeared to be this, that in the case of the new curves the iron was systematically 

 demagnetised by decreasing reversals of the longitudinal field, but in the case of the old 

 curves by decreasing reversals of the circular field. It was found, as a matter of 

 experiment, that the differences in the initial values of the shielding ratios depended on 

 whether the shields had been demagnetised by the circular or by the longitudinal 

 magnetising forces. This fact formed the starting point of experiments being made on 

 the magnetic aeolotropy of demagnetised iron, which have been incorporated in this 

 paper, together with others dealing with superposed magnetic inductions in iron, 

 §§ 14 to 37, to which the reader is referred. 



§ 55. Comparison of figs, xxviii. and xxix. with figs. xx. and xxi. at once shows that 

 the same general relationship exists inter se between the four curves resulting from the 

 four methods of field superposition in both cases. Also that the maximum value for 

 shield A of, say, the TL curve is at least approximately the same as the maximum value 

 of the TC curve (fig. xx.) ; but that the maximum value for shield B of the TL curve is 

 distinctly less than the maximum value of the TC curve (fig. xix.). When shield B, 

 however, was demagnetised by decreasing reversals of the circular field, not only was 

 the initial value of the shielding ratio raised, but the new curve assumed higher values 

 as it receded from the vertical axis. This is shown by the dot and dash curve TL, 

 fig. xxix., plotted from the last three columns of Table XXL, and its maximum value is 

 seen to be practically the same as the TC curve of fig. xxi. Experiments were not made 

 with shield A under the same conditions of demagnetisation. Table XXL, however, 

 supplies for shield A the shielding ratios obtained under the TL conditions for additional 

 values of the transverse field, viz., H t = 33'6 and 64*25 C.G.S units. The former 

 practically coincides with that already plotted when 11^=20 "9 units. The latter, 

 however, is plotted in fig. xxviii., which shows distinctly lower values of the 





