8 Mr. Barlow on the Efficiency of the Galvanometer 



On examining the errors arising from the assumption that 

 the deflection varies inversely as the square root of the length, 

 we find them amount on an average to about a degree, the 

 maximum exceeding 2° ; and Mr. Christie properly observes, 

 that with such errors we cannot admit the law of the square 

 roots. These errors however, employing the law of the in- 

 verse of the length, would amount to 18°; and it is, therefore, 

 very desirable to trace the cause of the discrepancy to its 

 source. 



There can be no doubt that M. BecquerePs experiments 

 were carefully conducted, and the means which I employed are 

 stated above; there ought therefore to be found some primary 

 cause for the disagreement, which is, after all, I suspect, not 

 very remote. In order to lead the way to such an investiga- 

 tion, I shall offer a few suggestions which have occurred to me, 

 leaving the bearing they may have on the subject to the consi- 

 deration of others. 



In the first place, I doubt much whether the deflection of a 

 compass needle, either from the action of a single wire or from 



* The constants employed in these computations have been obtained by 

 finding the value of all the tan A \f L's and taking the mean, and the 

 value of all the tan A. L's, and taking the mean of these. 



