546 PROFESSOR KNOTT ON THE STRAINS PRODUCED IN 



unquestionably an after-effect. In other cases the same peculiarity was observed, the 

 general character of which may be described as follows : Take any tube, such as Iron 

 VI. or VII. or Steel V. or VI., which shows diminution of volume in the lowest fields, and 

 subject it to a cycle of moderate range. Then every time the field, after having been 

 applied in one direction and removed, is applied in the opposite direction, there occurs 

 as a first effect an increase of volume. When the tube is subjected to repeated applica- 

 tions of the same field without alternation of direction, this initial positive effect is 

 frequently absent altogether, and in all cases is distinctly modified. All such 

 phenomena are evidently illustrations of the lingering effect of pre-existing magnetic 

 distributions ; but it is very difficult to understand their full significance. 



The cyclic graph of Nickel Tube V. is of the same type as the others, though com- 

 paratively simple. The chief difference is that there is no minimum, not even a hint 

 of a minimum. The return branch lies lower than the ascending branch of the curve, 

 and all the branches have maximum points. The line drawn half-way between the 

 ascending and return branches agrees very well with the corresponding curve on 

 Plate V. 



As noted above, the initial excursions at make depend in great measure upon the 

 fact of reversal. For example, when Steel No. 3 with brass cap was treated cyclically 

 by the alternating field ± 150, the volume change was + 1*06, and the initial excur- 

 sion + 3*27. But when the cycle was established by making and breaking the field 

 150 always in the same direction, the volume change was + '65, and the initial excur- 

 sion — '33. In other cases the sign of the excursion remained positive, but its magni- 

 tude was greatly diminished. 



A striking feature associated with the magnetic-strain hysteresis here discussed is 

 the remarkable effect of inertia in causing the tube, when subjected to rapid alterna 

 tions of field, to oscillate far beyond the limit of statical strain. The case seems to 

 be exactly analogous to that of a spring to which a load is suddenly attached. The 

 spring extends nearly twice as far as the extension which would be produced if the 

 load were applied statically. 



§ 13. Behaviour in Fields of Various Distribution. — The results discussed in 

 § 11 show that the effects we are studying are profoundly influenced by differences in 

 composition or in heterogeneity so slight as to have little appreciable influence on other 

 physical qualities. Two iron or steel tubes, cut from the same bar and bored out by the 

 same process and to the same extent, seem as diverse as possible in their behaviour in 

 the magnetic field. Increase of volume of bore is the prevailing characteristic of the 

 one ; decrease of the other. 



But it must be remembered that the measurement of the volume change of the 

 strain throws no light upon the change of form, which is no doubt an important feature 

 of the strain. The character of the strain as a whole we should expect to be very 

 complicated. For we know that the elongation of a thin rod uniformly magnetised is 

 not simply related to the magnetising force ; how much less simple will the relation 





