OP ARTS AND SCIENCES. 191 



XI. 



NOTES ON MAGNETIC DISTRIBUTION. 



By Henry A. Rowland. 



Presented, June 9, 1875. 



In two papers which have recently appeared on this subject, by Mr. 

 Sears (Amer. Jour, of Science, July, 1874), and Mr. Jacques (Pres. 

 Amer. Acad, of Sciences, 1875, n. 445), a method is used for deter- 

 mining magnetic distribution, founded on induced currents, in which 

 results contrary to those published by M. Jamin have been found. It 

 does not seem to have been noticed that the method then used does 

 not give what we ordinarily mean by magnetic distribution. In mathe- 

 matical language, they have measured the surface integral of magnetic 

 induction across the section of the bar instead of along a given length 

 of its surface.* M. Jamin's metnod gives a result depending on the 

 so-called surface density of the magnetism, which is nearly proportional 

 to the surface integral of the magnetic induction along a given length 

 of the bar. Hence the discrepancy between the different results. 



Had the expeiiments of Mr. Sears and Mr. Jacques been made by 

 sliding the helix inch by inch along the bars, their results would have 

 confirmed those of M. Jamin. Four or five years ago, I made a large 

 number of experiments in this way, which I am now rewriting for pub- 

 lication, and where the whole matter will be made clear. At present, I 

 will give the following method of converting one into the other. 



Let Q be the surface integral of magnetic induction across the section 



of the rod, and let Qe be that along one inch of the I'od : then Qe cc -j- 



X being the distance along the road. Hence, M. Jamin's results 

 depend on the rate of variation of the magnetization of the rod, while 

 those of Mr. Sears and Mr. Jacques depend on the magnetization. 



In conclusion, let me heartily agree with Mr. Jacques's remarks 

 about M. Jamin's conclusions from his experiments. Such experi- 



* Meavill's Electricity and Magnetism. Art. 402. 



