1894 - 95 .] Prof. Knott and A. Shand on Magnetic Strains. 335 
In the case of the nickel tube, the difference in the two columns 
is not of any great moment. With the nickel cap or stopper the 
changes of volume are slightly greater. This is very much what 
one would naturally expect to find. For the nickel stopper has 
the effect of lengthening by an inch or so the magnetisable bar, and 
the tube portion of the bar so lengthened will be more effectively 
magnetised than when the stopper is of a non-magnetic metal. 
But it will be noticed that the behaviour of the steel tube is 
greatly altered when the brass cap is substituted for the iron cap. 
Thus there is no change of sign in field 200 when the brass cap is used, 
and there is no maximum negative dilatation in field 400 when the 
iron cap is used. In short, the character of the stopper has a marked 
effect upon the whole character of the strain. It is easy to see that 
the iron stopper will tend to increase the induction into the tube ; but, 
except in the way of diminishing the induction, and so altering the 
accompanying change of length, we cannot credit the brass stopper 
with any effect upon the longitudinal strain. It is as a disturbing 
factor in the transverse strain that we must consider it. It does 
seem extraordinary, however, that a stopper, whose hampering 
action is confined to one end of a fairly long tube, should, by its 
inability to yield to the magnetising forces acting on it, produce 
such a marked effect upon the volume changes taking place in the 
tube as a whole. A closer investigation of the phenomenon may 
throw some light on the character of the strain accompanying 
magnetisation. 
