LABOliATOKY NOTES. 



289 



Table I. 



Since the second differences have appreciably ditferent values, it 

 i> iinpcjssihle to represent the law oi" chani^'e by means of a paralxjiic 

 liinction. ]>ut the reniarkaljle ('oustancy oï the ratios of successive 

 pairs of resistances sugi^'ests an exptjnential function of the temperature 

 as the expression for the resistsiuce. 



Thus we may put 



it. — il,, a 

 from which we tind, if / is the temperature in degrees Centigrade, 



k = -OOi^GOo, R, = -09519 

 The measured resistance at 7°*5 C. was 0*09604, which doe.> not diilcr 

 ivom the value given by the f)i'nuil;i 1)V more than 1 ])er cent. 



In my ])aper already referred t(j \ fxuid that the same form of 

 ex})ression held for the case of <jne of the nickel wires investigated, the 

 only essential difference Ijcing in the vahie of the coefïicient k, which 

 for the nickel was .003. 



The resistance of cobalt therefore d«)es not change so quickly with 

 temperature as does the resistance of nickel. 



Tn the second series of experiments the lower ends of the rods 

 with their connecting wires were insei-ted into a porcelain vessel. 

 Asbestos was wra])ped r<nmd the wires ; and the whole was heated in 



