308 
Proceedings of the Royal Society of Edinburgh. [Sess. 
molecules in those lower zones (equatorial) to obtain on summation a 
transverse effect corresponding to the experimental results. But this 
increase in the number of molecules considered would also increase the 
theoretical ratio between the co-directional and transverse effects, a nearer 
approximation to the experimental ratio B r /B L being thus obtained. The 
co-ordination of the transverse and co-directional inductions at all stages 
of the subsequent magnetising force must also be considered. 
Meanwhile it may be concluded that the experimental results can be 
co-ordinated with a preponderance of molecules within a broad equatorial 
belt in reference to the demagnetising (polar) force. 
iEOLOTROPY. 
In the paper on the “ Molecular Condition of Iron demagnetised by 
Various Methods,” a comparison was instituted between the changes 
involved in the process by which iron passes from the condition in which 
it is left demagnetised by decreasing reversals and by annealing, as 
residual magnetisation is superposed upon both these conditions. By 
putting on and withdrawing, say, a positive field the iron was left residually 
magnetised, but in a condition more susceptible to negative than positive 
change. This simple method introduced a more complicated form of 
magnetic seolotropy than that which results from demagnetisation by 
decreasing reversals. 
In the latter case the iron is isotropic in reference to the positive or 
negative application of the same directional force producing the de- 
magnetisation, but aeolotropic in all other directions. 
In the former case the iron is molotropic to subsequent positive and 
negative applications of the same directional force producing the residual 
magnetisation, and it is also seolotropic in all other directions. A condition 
of this kind does not satisfy even approximately one of the theoretical 
assumptions made — that on the whole the molecules rotate along great 
circles of longitude, the initial direction of molecular rotation depending 
largely upon the previous history of the iron.* 
In this investigation therefore a method of partial demagnetisation by 
decreasing reversals is adopted. The steel tube is in the first place subjected, 
by means of the revolving commutator, to reversals of the maximum 
demagnetising force. On the commutator being stopped, the maximum 
value of H is left positive. The field is withdrawn and reversed to a 
smaller negative value of, say, — H r Demagnetisation from this maximum 
* “The Shift of the Neutral Points, etc.,” Proc. Roy. See. Edin ., vol. xxix. pp. 30-34. 
