Method of Galvanic Calibration of a Wire. 61 



the calibration of a thermometer. It starts from the same 

 idea on which the well-known Hockin-Matthiessen method of 

 employing the bridge rests. 



In the annexed figure let A N B and A M B be the two 

 branches of the current in the known bridge-arrangement — 

 A M B the wire to be calibrated, A N B a series of resist- 

 ances, whose sum with any arrangement remains constant. 



Now let M 1 and N l5 or M 2 and N 2 , be two pairs of points of 

 equal potential, and let y denote the resistance N x N 2 , x the 

 section M 1 M 2 of the wire ; then 



y=Qx, 



where Cis a constant (the " constant of sensitiveness ") which 

 depends only on the sum of the resistances ANB. In fact, if 

 W denotes the sum, and L the total length of the wire, 



W=C.L. 



Further, let a be the calibration-interval, so chosen that 



- = n is a whole number. 

 a 



Now, first of all, n nearly equal resistances are prepared. 

 We had a number of tenths of a Siemens unit already at hand, 

 which we had previously procured and could employ for the 

 purpose of resistance-determinations. A German-silver wire, 

 however, of suitable length and thickness can be taken and 

 cut into n approximately equal parts, which it is then expe- 

 dient to solder to pieces of amalgamated copper wire. Exact 

 equalization is superfluous. 



These n approximately equal resistances I, II, III, &c are 

 placed in a series one after another by the interposition of 

 mercury-cups. We will assume w = 5. The above figure 

 represents the arrangement for this case. The connexion 

 with the extremities A and B of the wire is effected by means 



