656 MR JAMES RUSSELL ON 



§ 41. As already stated, the shielding effects plotted in figs. xx. and xxi. were 

 obtained with the transverse field equal to 20'9 C.G.S. units. Table XV. supplies the 

 data for shield A and Table XVI. for shield B, when the transverse field = 4 '37, 20*9, 

 33"6 and 64'25 C.G.S. units. With the weakest transverse field, viz., H f = 4 , 37, it was 

 found necessary to eliminate the disturbing effect of the leakage field (R) p within the 

 shield, due to the overlapped joint, when circularly magnetised. By the rotation of the 

 inductor, this field could be measured, the overlapped joint and the coil of the inductor 

 being in the same plane. It was found to amount to nearly 1 C.G.S. unit when the in- 

 duction in the iron was B= 12,500. By rotating the shield 90° and making the final 

 adjustments in azimuth by experiment, the effect of this leakage field when the inductor 

 was rotated could be reduced to a minimum. This adjustment was made for both 

 shields before the measurements tabulated in Tables XV. and XVI. were made. In 

 addition to this, each set of measurements (H c being increased from zero to a maximum) 

 were repeated twice, with the circular magnetising force first in one direction and second 

 in the opposite direction, the average readings being given in the second columns (H) TO . 



§ 42. Table XIV. gives in full detail the various sequences of measurements adopted 

 under the conditions of superposition of transverse and circular fields applicable to 

 Tables XV. and XVI. The measurements under the TC and TCC conditions do not 

 appear to call for any remark ; but, under the CT and CTT conditions, attention may 

 be called to the necessity of demagnetising the shields after the readings taken with 

 each + or — increment of the circular field, in order that the superposition of the 

 transverse field upon increments of H c from zero (magnetisation) might be maintained 

 throughout. The experimental data given in Table XV. (shield A) and Table XVI. 

 (shield B) are plotted as shielding ratio curves (with the exception of the TCC curves) 

 for the four selected values of the transverse field, viz., H c = 4 - 37. 20"9, 33' and 64"2o 

 C.G.S. units. 



under the TC conditions, 

 under the CT conditions, 

 under the CTT conditions. 



In 



fig- 



XXII. 



for Shield A 



JJ 



fig. 



XXIII. 



» » B 



In 



fig- 



XXIV. 



for Shield A 



i) 



fig. 



XXV. 



„ „ B 



In 



fig- 



XXVI. 



for Shield A 



?» 



fig. 



XXVII. 



„ „ B 



Inspection of these curves shows that for each condition of superposition of fields, the 

 differences which exist in their initial values, due to the transverse field alone, diminish 

 usually at first and always finally as the circular magnetising force increases ; and that 

 the curves given in figs. xx. and xxi., when H c =20'9, may be regarded as typical so 

 long as the values of the transverse field are not too high. The differences which exist 

 between the four types of curves corresponding to the TC, TCC, CT, and CTT 

 conditions are sufficiently evident, and are a maximum when the values of dB/dH due 

 to the circular force are a maximum for shields A and B (see fig. iv.). Thus the 

 maximum shielding ratios when the circular field is simply superimposed upon the 



