62 On Galvanic Calibration of a Wire. 



of thick copper wire or plates. The calibration now takes 

 place in the following manner: — 



The contact N of the " bridge-wire " M N, passing through 

 a sensitive mirror-galvanoscope, is put successively into the 

 cups 1 and 2, and the corresponding positions M, and M 2 , 

 upon the wire A M B, of the contact are determined. 



I and II are now exchanged, so that I comes into the place 

 of II ; the contact N is put successively into the cups 2 and 

 3 ; and again the corresponding positions M' a M 3 of the con- 

 tact M are determined. We now move I further on, by making 

 it and III change places, put the contact N successively into 

 3 and 4, and determine the positions M' 3 and M 4 . 



We proceed in this way until I has arrived at the place of 

 the last resistance V. 



We thus obtain, upon the bridge, the lengths M x M 2 , M' 2 M 3 , 

 M' 3 M 4 , &c, which are each proportional to the same resistance. 



It is easily seen that, in fact, between this method of wire- 

 calibration and the well-known method of thermometer- cali- 

 bration a perfect analogy prevails. As in the latter a mer- 

 cury thread of a determined volume and the approximate 



length a— is shifted in the thermometer-tube, and in the 



graduation of the thermometer its length is read off, conse- 

 quently a quantity which, on the assumption of equal cross 

 section of the tube, is proportional to that volume, so also in 

 our method a wire possessing a determined resistance is 

 shifted from place to place, and is read off in the division of 

 the length, consequently likewise a quantity which, on the 

 assumption of equal cross section of the wire to be calibrated, 

 is proportional to that resistance. From the differences of 

 equivalent lengths we infer in both cases respecting the calibre 

 of the tube and the wire respectively. The fixed points between 

 which the divisions are put in are, in the case of the thermo- 

 meter, the freezing- and the boiling-point ; in our case they 

 are the two branching-points A and B. 



"Now, if %, « 2 , «3, &c are the equivalent wire-lengths read 

 off, then the mean length which, assuming a mean cross sec- 

 tion of the calibrated wire, corresponds to the calibration- 

 resistance is given by 



a x + a 2 + a s + . . . + a n 



As «!, a 2 , . . . deviate each but very little from the calibration- 

 interval a— -, it is more practical for the calculation to bring 



