162 Dr. E. H. Hall oh the "Rotational Coefficient " 



H. Freeman, Fellow in Physics, and Mr. H. K. Goodnow, 

 Special Student in Physics, who for a while carried on 

 the experiments together. Mr. Freeman especially worked 

 with me for a long time ; and several suggestions of his in 

 regard to the arrangement of apparatus and the method of 

 experimenting were adopted with great advantage to the work. 

 In my last article on this subject the results of measure- 



M x Y 



ments were given in the form — wr~} where M was the 



strength of the magnetic field*, V was the direct current 

 divided by the section of the conductor, and E / was the trans- 

 verse electromotive force per centimetre of the width of the 

 strip. In that article were given certain reasons for thinking 

 the above quantity more likely to be a constant for any given 



metal than the quantity — ™-,-- , where E is the electromotive 



force per centimetre of the length of the metal strip. Recent 

 developments, to be spoken of further on, raise the question 



whether the ratio —. ™ — will not after all prove to be the more 



fundamental and invariable quantity ; but as E is rather dif- 

 ficult to determine with accuracy, and as in any given strip of 

 metal Y is likely to remain under ordinary conditions of tem- 

 perature &c. very nearly proportional to E, the use of the 

 former quantity will be retained for this article at least. The 

 values of M [§] will be given separately, however; and, for 



F' 

 convenience in plotting the results, the quantity ^ will be 



used instead of ^7. The values of ^ will, in plotting, be laid 



E' 



off as abscissas, and the values of -^ be taken as ordinates. 



This method of plotting gives a simple curve in the present 



case, and puts the results of the experiments in form to be 



compared with those of previous investigations of some of the 



E' 

 magnetic properties of nickel. It is this quantity ^ which, 



after Maxwell, in accordance with the suggestion of Mr. Hop- 

 kinsont, is now called the " rotational coefficient" of nickel. 



* Called $ in this article. 



t Phil. Mag. Dec. 1880, p. 430. Prof. Rowland has (Phil. Mag. 

 April 1881, p. 254) remarked upon Mr. Hopkinson's note. Maxwell did 

 not know any such effect to exist. In fact he expressly stated that it pro- 

 bably did not exist ; yet, seeing the possibility of it, he let fall the phrase 

 which seems now best fitted to define this newly discovered property of 

 the metals. 



