STEEL, NICKEL, AND COBALT TUBES IN THE MAGNETIC FIELD. 471 



measured volume changes of bore of the Iron Tubes I. to VII. (see Plates I. and II. of 

 the previous paper) differ markedly from those of the Iron Tubes F. to VIII'., and 3, 5, 

 and 7. Nevertheless, the dilatations of Tube VII., as given in Table VIII. below, do not 

 appreciably differ from those of other tubes. 



It is this consideration which fully explains the apparently extraordinary behaviour 

 of the steel tubes as described in the first paper (p. 537 and Plates III. and IV., I.e.). 

 I have no measurement of the volume change of steel, so that it is not possible to 

 calculate even the tentative values of v, n', and v'. Probably 8 is small compared with A, 

 and some idea of the character of the strain in steel might be got by assuming S to be 

 negligible. The results would not, however, differ essentially from those for iron. I 

 have given in Table X. simply the undoubted results for the Steel Tubes, 3, 5, 7, VII. u 

 VII. 2) VIL 3 , and 9. There are three sets of columns of numbers. The first contains the 

 ratios X + 2m calculated mainly from the data given in Part I. ; the second contains the 

 measured values of X ; and the third contains the calculated values of m. Broadly 

 speaking, they are very similar to the corresponding values for iron. 



The history of Steel No. VII., as given in the Appendix to Part I., is very 

 extraordinary. In its earliest condition distinguished as VII.! after the final boring 

 which changed VI. into VII., it behaved, as regards bore dilatation, in a manner 

 altogether peculiar. There was great positive dilatation up to Field 280, and negative 

 dilatation in higher fields. Two months later, the law of the dilatation was found to be 

 just reversed. This condition is distinguished as VII. 2 . The tube was then annealed by 

 slow cooling ; and the bore dilatations were found to be greatly diminished in value, 

 but otherwise to resemble roughly the corresponding quantities in the second 

 condition. The contrasts between these three successive states of the same tube are 

 shown in the numbers in Table X. The transverse dilatations m for Steel VII. 1 and 

 Steel VII. 2 are calculated on the assumption that the longitudinal dilatation of each is 

 the same as that for VII. 3 . A consideration of the three cases shows very plainly the 

 effect of comparatively small changes in the values of the tangential dilatations. In 

 none of them does 2/x differ greatly from — X. Thus the ratio X-f 2/x is comparatively 

 small in all. In VII. 1, X has the preponderating influence ; in VII. 2 , 2/u. preponderates. 

 In VII.3, 2u still preponderates, but very slightly ; so that in this case the bore 

 dilatations are very insignificant compared to the linear dilatations. 



§ 9. The Coiled Iron Tubes C I., C II. (Table V.) — These were formed of sheet-iron *37 

 mm. in thickness. They were each about 10 inches long. The diameter of C I. was shorter 

 than 1 inch by 4 per cent., and the diameter of C II. longer by about the same fraction. 



No appreciable change of volume of the metal was obtained, but this simply means 

 that it was too small to be measured. Under the most favourable conditions of 

 experiment half a small division of the micrometer scale was about the limit of certainty. 

 This would mean a volume change of about 2 x 10 " 6 cub. cm., giving a cubical dilatation 

 of *25 x 10 " 6 . I believe half of this value might have been detected. We may safely 

 conclude that the cubical dilatation of this kind of iron, when placed in a magnetic field 



