1897-98.] Prof. C. G. Ivnott on Strains in Iron, etc. 217 
very small and practically negligible value of the cubical dilatation, 
fx has the value + 18 x 10 -6 . 
With nickel tubes formed by coiling thin plates, A. reaches the 
value - 30 x 10 -6 , while [x is found to be + 18 x 10 -6 , v being 
consequently + 12 x 10 -6 . This excess in value of fx over v in very 
thin- walled tubes, and the excess in value of v over jx in tubes of 
thicker walls, are points of some interest. 
In iron and steel, X is positive in low and moderate fields, and 
negative in high fields. The ratios /x and v tend to have opposite- 
signs from X ; but the tendency is considerably modified in the 
tubes of narrower bore. As the bore is increased and the thick- 
ness of wall diminished, the well-marked maximum in X occurs in 
a lower field. The cubical dilatation is always positive. The 
following values for iron tube BY will show the nature of 
the changes which occur in these ratios as the field is gradually 
increased. 
Elongations x 10 6 . 
Field 
At Inner Surface. 
At Outer Surface. 
A 
V 
A' 
r 
V 
190 
+ 2-1 
-1‘5 
- -4 
+ 2-1 
-1-25 
- -65 
310 
0 
- *4 
+ *6 
0 
- *1 
+ *3 
500 
-2-2 
+ -8 
+ 1-6 
-2-2 
+ 1-0 
+ 1*3 
Yery similar values are found with the other tubes discussed. 
Among other general conclusions arrived at the following may 
be stated 
The apparently capricious character of the results obtained in 
Part I. finds a ready explanation. The bore dilatation ( X + 2/x) 
appears as a differential effect, being the sum of two quantities of, 
in general , different signs ; and of these, sometimes the one and 
sometimes the other has the greater value. 
Except in the case of very thin walls, a circular element in any 
transverse plane, perpendicular to the axis of the cylinder, be- 
comes, when the cylinder or tube is magnetized parallel to its 
VOL. XXII. 3/10/98 P 
