604 
PROFESSOR J. A. EWING ON EXPERIMENTAL 
(1) Demagnetisation by reversals with no load on. 
(2) Application of the magnetising field with no load on. 
(3) Loading and unloading (0—15 — 0) until a nearly cyclic state was reached. 
The effects of the first loading in each field are shown by dotted lines in Plate 63, 
fig. 41, and the cyclic effects reached after several repetitions of the loading are 
shown by full lines in the same figure. 
Thus the point a (fig. 41) is the magnetism reached by applying a force 0'34 with¬ 
out load. The curves ab, be are the effects of loading to 15 kilos, and unloading in 
this field. The full lines just above them are the still very imperfectly cyclic effects 
given by a repetition of the cycle 0 — 15 — 0 in the same field. 
Similarly the points d and e are the values of magnetism reached by applying the 
fields 2‘49 and 18'65 respectively; the dotted lines starting from these points show 
the effects of the first loading in those fields ; and the full lines above them the cyclic 
effects after two or three loadings. 
The curves fg h show the effects of stress on the residual magnetism left after the 
removal of the field 2'49. The wire has been more stretched here than in the experi¬ 
ment of § 86, and, probably for that reason, appears to be somewhat more persistently 
retentive in the present case. 
It is interesting to notice here how the susceptibility to the action of stress, as 
regards both the first general augmentation and the subsequent cyclic changes of 
magnetism, is greater in the intermediate field of 2'49, than in either the strong 
or the weak field. In the weak field the magnetic effects are still not cyclic after 
many repetitions of the cycle of load. In the high field, on the other hand, a 
sensibly cyclic state is reached after a single loading. 
The maximum point occurs at a lower load as the strength of the field is raised. 
The general effect of putting on 15 ki]os. is positive for the low field, more 
strongly positive for the intermediate field, but negative for the high field. This 
reversal of the effects of on and off, by increase of the magnetising field, was noticed 
first by Ahllari, and afterwards independently by Thomson, who has since named it 
the “ Villari reversal.”* 
§ 88. To study more particularly the forms of the curves of magnetism and load 
during the passage of the effects of stress through a maximum, and the reversal 
of these effects in high fields, the following series of observations was made, in which 
a variety of magnetising fields ranging up to 34 e.g.s. units were employed. 
(May 8th, 1882.) The same wire was further stretched under a load of 18'5 kilos., 
suffering an additional elongation of 3 per cent. After demagnetisation by reversals, 
with no load on, a magnetising field was gradually applied, and then kept constant 
during the changes of load. Loads were applied, up to 18'5 kilos., removed, 
reapplied, and so on. The load was then removed, the wire demagnetised by 
* Villari : Pogg. Ann., CXXVI., 1868; Thomson: Phil. Trans., 1879, p. 55. 
