594 Proceedings of Royal Society of Edinburgh. [sess. 
II. The effects of hysteresis will be to produce a magnetic 
seolotropy, the main character of which is not difficult to imagine 
or describe. 
Thus let H' be the circularly magnetizing force acting on a 
molecule of the tube, and let the longitudinally magnetizing force 
H be applied first in one direction, then in the opposite direction 
( + H, - H). Were there no hysteresis, the resultant twist would 
be as if the principal elongations were first in the direction of one 
of the dotted diagonals, and then in the direction of the other. 
But because of the persistence of IT, conjoined with the effects of 
hysteresis, the direction of principal elongation as the longitudinal 
Fig. 2. 
field changes cyclically from + H to - H and back again to + H 
will oscillate between the double arrow-headed lines and will be 
less inclined to H' than the dotted diagonals (see fig. 2, a). In 
other words, the result is as if a more powerful H' acted without 
hysteresis ; but this means in general a greater twist, since H > H' 
usually, a result in full accord with (1). 
On the other hand, if we take the case in which the longitudinal 
field (H) is constant, while the circular field (H') is varied 
cyclically from + H' to - H' and back again to + H', we see at 
once that the direction of principal elongation will oscillate 
between the double arrow-headed lines less inclined to H than 
the dotted diagonals. When H is greater than H', this gives a 
