406 



Prof. D. E. Hughes. 



[Mar. 6, 



Before commencing my researches upon neutrality, I felt that it 

 was necessary to observe the curves of magnetic penetration, whilst 

 under the influence of its inducing cause. It is well known, however, 

 from the researches of Graugain, Du Moncel, and Jamin, that the 

 magnetism does not penetrate to a very great depth with its full 

 force, decreasing rapidly from the exterior to the interior. Most 

 observations have been made by means of tubes of various thickness, 

 introduced into each other. These, however, introduce an element of 

 error, as, in separating them, they are necessarily drawn over each 

 other. Jamin's method of dissolving the exterior of a steel magnet 

 in diluted sulphuric acid gave results free from experimental error , 

 but this could only be employed after the cessation of the inducing 

 cause, the observations being really upon the permanent remaining 

 magnetism. 



The methods employed by myself consisted, first, in superposing- 

 twenty flat iron strips, \ millim. thick, 20 centims. long, and 3 centims. 

 in width. These could be built up into a solid rod 1 centim. total 

 thickness. Each piece was carefully selected and measured for its 

 magnetic capacity, so that they should all be equal in value whilst 

 under the influence of an inducing force, as well as their remaining- 

 magnetism when the influence ceased; the remaining magnetism 

 being about J of its capacity in the size and kind of iron employed. 



These strips forming a compound bar were placed in contact with 

 the poles of a strong permanent magnet, or they could be laid on one- 

 pole, the object being to polarise the lower bar only by contact, and 

 observe the degree of penetration. The upper strip was carefully 

 separated whilst the remainder was left under the polarising influence. 

 We could thus separate each bar while under the influence without 

 fear of reactions taking place between the separated bar and its- 

 companions. We had thus a bar or strip, separated while under the 

 inducing influence, and, knowing its coefficient of remaining mag- 

 netism, we could estimate its full power when under the polarising 

 influence. By this means the values were plotted graphically, 

 giving curves of varying degrees, as the inducing force was changed, 

 or the material of the strip from soft iron to hard steel. These 

 curves were verified by a somewhat similar method, using a separate- 

 strip whose coefficient of remaining magnetism was known, and 

 drawing this over the poles of a magnet, but separated from it by 

 different degrees of thickness of iron. 



These again were verified by an electro-magnetic method, in which 

 a series of concentric tubes divided lengthwise was employed, so as to 

 allow separation without friction, confirming the numerous curves- 

 obtained by the previous methods, showing that with a limited magne- 

 tising power acting upon homogeneous iron or steel, the penetration is 

 inversely as the square of the distance from the inducing power, but 



