1890.] 



on Magnetic Phenomena. 



51 



traversed by perhaps as many as 18,000 lines per centimetre. Although 

 this is a very large number, it will be noticed that it is smaller, in 

 proportion, than was obtained when the magnetic field of the earth 

 alone was employed. In that case a field of half a line to the centi- 

 metre was found to induce 300 lines in the iron, the multiplying power 

 being 600. But with an external field of 200 the multiplying power 

 is only about 90, a very considerable falling ofi". It is usual to 

 denote the number of lines per square centimetre in the magnetic field 

 by the letter H,and those induced in the iron by B, while the multi- 

 plier is indicated by the Greek letter /x. We may therefore write : — 



B - /xH. 



B is commonly spoken of as the " magnetic induction," and fx as the 

 " permeability." 



It used to be assumed that, except in strong fields, the permeability 

 fjL was practically a constant for the same specimen of metal. We 

 have already seen that this is by no means the case, and how very 

 far it is from being so is clearly shown by the following table, in the 

 first and third columns of which are given corresponding values of 

 H and fjL for an average specimen of wrought iron. 



Table I. — Iron. 



It will be remarked that, as the strength of the field increases 

 from the smallest values, the permeability at first rises w^ith enormous 

 rapidity, attaining in a field of 2 or 3 units a maximum value of more 

 than 2000 ; then it falls again, rapidly at first, and afterwards more 

 slowly, until with a field of 65 lines to the centimetre the permeability 

 is no more than 255. So far, the figures in the table (which are given 

 in round numbers) are \jasedupon experiments made by Prof. Eowland 

 sixteen years ago. Plotting corresponding values of /x and B, Rowland 

 constructed a curve, the form of which led him to the remarkable 

 conclusion tha£ the value of the magnetic induction B could not 



E 2 



