STEEL, NICKEL, AND COBALT TUBES IN THE MAGNETIC FIELD. 



465 



that the maximum value of X in each A tube occurs in a lower field than it does 

 in the corresponding B tube. Thus the field of maximum A is : — 



In A I. 165 and 370 in B I. 



„ A II. 



150 



)) 



310 



53 



B 11. 



„ A III. 



140 



55 



270 



33 



B III. 



„ A IV. 



125 



55 



225 



33 



B IV. 



„ A V. 



110 



35 



190 



33 



B V. 



„ A VI. 



95 



3> 



150 



35 



B VI. 



„ A VII. 



80 



35 



100 



55 



B VII. 



Thus the B tubes reproduce, but in higher fields, the peculiarities shown by 

 the A tubes. The reason is not far to seek. It depends on the fact that the 

 shorter B tubes have a larger demagnetizing factor than the larger A tubes. Not 

 only so, but, in accordance with well-known results, the demagnetizing factor 

 diminishes with the area of section of the material, the length being constant. 

 This consideration explains at once the gradual shifting of the critical points 

 (maximum A, minimum Sv and St/) into lower fields as the bore of a tube of given 

 length is gradually increased. 



The shifting of the crest of the longitudinal elongation curve as we pass through 

 the series A I. to A VII. and B I. to B VII. is also well shown with tubes F. to VIIF. 

 An inspection of Table IX. will bring this out clearly enough ; but the feature is 

 most distinctly shown in the following table constructed from the curves corre- 

 sponding to Table IX., which curves, however, being broadly similar to those in 

 Plate II., I have not thought it necessary to publish. 



Fields for Maximum Elongation in Iron Tubes I', to VIII'. 



Tube. 



I'. 



II'. 



III'. 



IV. 



V. 



VI'. 



vir. 



vnr. 



Field 



230 



215 



200 



180 



170 



155 



140 



120 



According to the commonly accepted theory, the demagnetizing factor in a 

 cylindrical bar is proportional to the square of the ratio of the diameter to the 

 length. In the case of a cylindrical tube this law requires modification. Perhaps 

 the most probable simple hypothesis is to compare the tube, as regards its 

 demagnetizing factor, to a bar of the same length and the same cross-section of 

 material. How far this applies to the present case is tested immediately by a 

 comparison of Tubes A II. and B VI., which have their maximum elongations in 

 about the same field. Presumably their demagnetizing factors are nearly the same. 

 Now the cross-sections of the material of A II. and B VI. are as 5 "96 to 1*94. 

 Dividing these respectively by 4 and 1, which are as the squares of the lengths of 

 the A and B tubes, we get for the ratio of the demagnetizing factors 194 : 144 or 



VOL. XXXIX. PART II. (NO. 15). 4 A 



