MAGNETIC SHIELDING IN HOLLOW IRON CYLINDERS. 637 



I. (a) Experimental Shielding Ratios. — Transverse Field increased by 



increments from Zero. 



§ 10. Table V. gives the experimental data under the first above-mentioned conditions, 

 viz., with increasing values of the transverse field H t . The iron shields having been 

 previously demagnetised, H t was increased from about 4 C.G.S. units to a maximum of 

 130 C.G.S. units. At each step H t was repeatedly reversed in order to secure a stable 

 magnetic condition, and one which could be repeated with a minimum of error. The 

 first column gives the values of the transverse field, the second column the corresponding 

 values of the weakened field within the shield obtained by the rotation of the inductor, 

 and the third column the experimental shielding ratio g' [H,/(H)J. In fig. vi. these 

 ratios for shields A and B are plotted as ordinates against the corresponding values of YL t . 



These curves now fall to be compared with the theoretical shielding ratios. The 

 induction round the sides of the shields is not uniform (see fig. xxxiv. (1)), and 

 obviously reaches a maximum on both sides of the shields, where the iron is cut 

 by a plane which is normal to the transverse field, and which contains the axis of 

 the cylindrical shield. The induction at one of these positions, m, was measured by 

 means of the exploring coil shown in fig. in., and is plotted in fig. v. (full line 

 curves) from columns 1 and 2 of Table III. When the induction is carried beyond the 

 initial stages, Du Bois remarks that " the difficulty is to assign the corresponding mean 

 value of the true magnetising force, which is less than the impressed and greater than 

 the shielded field, though probably nearer the latter." * No difficulty, however, exists in 

 assigning values for each of the theoretical ratios (g) and ((g)) corresponding to the 

 induction at the maximum positions indicated, and it remains to be seen how far either 

 of these ratios approximates to the experimental shielding ratios (fig. vi.). By taking a 

 sufficient number of points on the induction curves (fig. v.), the same values of 

 induction in the B-li curves of fig. iv., where II is the true magnetising force in the 

 iron, supply on the same vertical ordinates corresponding values of the theoretical ratios 

 (g) and ((g)) from the dash and dotted curves respectively. These values will be found 

 in the third and fourth columns of Table III., and are plotted against the corresponding 

 values of the transverse field in fig. v. 



§ 11. It is at once evident that the ((g)) curves (dotted lines) reach considerably 

 higher values in the earlier stages of induction, and lower values in the latter stages of 

 induction, when compared with the experimental curves of fig. vi. On the other hand, 

 the (g) curves (dash lines) are very similar to those of fig. vi., attain, within the limits of 

 experimental error, the same maximum and minimum values, although, it must be noted, 

 •at somewhat lower values of the transverse field. But the theoretical shielding values 

 have been based upon the maximum measurements of induction round the shields, which 



* Du Bois, Electrician, "Magnetic Shielding," vol. xl. p. 815(1898). Fig. x. shows the relation between the 

 shielded field and the true magnetising force in the iron at the position of maximum induction. 



