employed for determining the Ohm. 259 



stants of the galvanic current, as fundamental units, and ex- 

 pressed a wish that a special international commission should 

 be intrusted, in the first place, with the construction of a 

 standard ohm as unit of resistance. Since, then, further 

 consultation is to take place before very long, it seems desi- 

 rable to consider again the methods hitherto employed, with 

 their sources of error, from the experimental standpoint ; their 

 mathematical theory has been sufficiently discussed. 



I desire also to give an impulse to further discussion of the 

 methods to be adopted in the new determination. This is to 

 be done the more strictly and thoroughly, since the units, 

 having been once determined, ought not to be altered again 

 immediately in consequence of further investigations. 



It is well known that TV. Weber, to whom we owe the fun- 

 damental facts in this subject, has given four methods of ob- 

 taining a conductor of given resistance in electromagnetic 

 units: — 



I. A circle of wire of known dimensions is caused to revolve 

 through a certain angle about an axis (vertical) inclireito 

 the direction of the earth's magnetism ; and the intensity of the 

 current thus induced is measured by means of a galvanometer 

 of known dimensions. The intensity of the current, under 

 similar conditions, is inversely proportional to the resistance 

 of the conductor. 



II. Instead of measuring the dimensions of the galvano- 

 meter in the first method, the action of the unit current on the 

 magnetic needle of a multiplier is determined by the damp- 

 ing of its oscillations when the circuit is closed. 



III. A magnetic needle is allowed to oscillate within a closed 

 multiplier of known dimensions, and the damping of its oscil- 

 lations is determined. 



IY. A circle of wire is put into uniform rotation about a 

 horizontal or vertical diameter; and we observe the deflection 

 which takes place in a magnetic needle swinging at the centre 

 of the circle, in consequence of the current induced in the 

 revolving circle by the earth's magnetism. Since in carrying 

 out these methods each separate measurement is of necessity 

 attended with error, that method appears at the outset the most 

 reliable which involves the determination of the fewest con- 

 stants, and in which these determinations can be made with 

 most accuracy. Hence methods III. and IV. appear at the 

 outset to offer special advantages. 



We will consider the fourth method first, in order to dis- 

 cuss at once a number of sources of error which partly affect 

 also the other methods. This method is in fact that employed 

 S2 



