498 PROFESSOR KNOTT ON SOME RELATIONS BETWEEN 



twist in nickel occurs in field /3 = 65, and its value is 0= —28 "3. The circular field 

 at the circumference of the wire is 



a-' 684 -14-5 



The greatest twist in the iron is 6= +6'8, occurring in field fi= 17 '5, and for a current 

 producing at the circumference a field 



•624 



•04 



= 156 



These values give 



- 28 " 3 = 7^15 x ^tk for ni ckel 

 047 4440 



+ 6'8 = -^-x y^ for iron. 

 04 550 



XT h 6-8 1 



Hence ^ = _ =— — 



Jc 141-5 21 



Now, according to Mr Bidwell's measurements, iron of length unity in a field of 21 

 units lengthens by 5 x 10" 7 , and nickel of length unity in a field of 67 units contracts 

 by 110x10" 7 . Thus the ratio of the elongations is — ^, being almost exactly the 

 same as the ratio of the multipliers just found. In other words, these multipliers for the 

 two different metals are proportional to the elongations. Thus Maxwell's explanation 

 receives a further corroboration. It applies equally well to nickel and iron ; so that, 

 although the formula used is admittedly based on assumptions only approximately 

 fulfilled, and at best cannot be expected to apply rigorously to wires, yet the approxima- 

 tion is almost exactly the same for both nickel and iron. When it is borne in mind 

 how different the strain effects are in these metals, it will surely be admitted that the 

 remarkable equality in the ratios of the twists and of the elongations proves a very close 

 connection to exist between them. 



9. Evidences of JEolotropy and After-effect. — One of the most interesting features 

 brought out by these experiments is the dependence of the twist on the order in which 

 the magnetising forces are applied. In other words, longitudinal and circular magnetis- 

 ing forces do not, so far as regards their strain effects, fulfil the simple law of superposi- 

 tion or composition. The elongations and twists involved are certainly within the limits 

 of elasticity ; hence we may conclude that these magnetic forces are not superposable as 

 regards their magnetic effects. Evidently, then, we are dealing with another instance of 

 the general phenomenon of the magnetic after-effect, or hysteresis, as Professor Ewing 

 has named it. 



The best known illustration of this phenomenon is the residual magnetism which 

 persists in iron or nickel after the magnetising force has been removed, or, as it is more 

 convenient to express it at present, after a given magnetising force + b has been 

 reduced to zero by the superposition of an equal but opposite force — h. In such 



