MAGNETIC SHIELDING IN HOLLOW IRON CYLINDERS. 659 



the total magnetic flux across the shielded space thus obtained plotted (dot and dash line 

 curves), to an ordinate scale on the right of figs xxix. and xxxi. respectively. These 

 scales having been made equal to the induction scales on the left, multiplied by d (for 

 shield A, d = "0233, and for shield B, d='029), it follows that all the curves of figs. 

 xxviii. to xxxi. measure total magnetic flux when referred to the ordinate scales on the 

 right. 



§ 45. Hence by means of the simple formula for N c given immediately above may 

 be derived from the full line curves of figs, xxvui. and xxx. (or, better still, from the 

 Tables XVIII. and XVII.), taken in conjunction with the dot and dash line curves of 

 these figs., the full line curves TC and CTT of fig. xxxii. (shield A), which will measure 

 the total number of circular magnetic lines wholly completed in and round the iron, 

 under the TC and CTT conditions respectively. Also, in a similar way, by means of the 

 formula for N t may be derived the dash line curves TC and CTT of fig. xxxn. (shield 

 A), which will measure the total number of magnetic lines which are continuous with 

 those of the transverse field, under the TC and CTT conditions respectively. Fig. 

 xxxni. gives the corresponding curves for shield B, obtained in the same way as those 

 of fig. xxxii. for shield A. In both cases the transverse field H, = 20 , 9 C.G.S. units. 

 The ordinate scales on the right have been retained, but it is apparent that the curves 

 can only correctly be referred to the flux scales on the left. 



§ 46. It would therefore appear that under the CTT conditions the wholly circular 

 induction is greater, and for low values of the circular force, very much greater, while 

 concurrently with this the magnetic lines continuous with the transverse field are some- 

 what less than under the TC conditions. When the transverse field is equal to 4*37 

 C.G.S. units, the induction curves for both sides of the shields A and B (dash line 

 curves of figs. xxx. and xxxi.) are, except near their origin, above the normal (dotted 

 line) curve when H t = 0. Reference to Table XVII. shows that this increase is brought 

 about by the creeping up of the induction upon each reversal of the transverse field, 

 superimposed upon a pre-existing induction to the circular field. 



The effect is evidently one of molecular vibration. The transverse field super- 

 imposed in one direction upon the pre-existing circular induction due to H c supplies 

 the first molecular tap which increases the induction in one side of the shield more 

 than it diminishes it on the other. When, however, H ( is repeatedly reversed, the 

 increase in the one side of the shield finally becomes equal to the decrease in the other 

 side, so that under these repeated molecular taps the induction becomes stable at con- 

 siderably increased values (subject to the induction not being pushed to extreme values 

 by either magnetising force). This result merely confirms that of other experimenters 

 working with cross magnetising force (H c and H t act in the same plane). Ewing, # 

 summarising a paper by G. G. Gerosa and G. Finzi, writes : " When a rapidly alternating 

 current of moderate strength traversed the wire, the susceptibility to longitudinal 



* Magnetic Inductions in Iron, 3rd edition, p. 330. 



