ON THE SPECIFIC RESISTANCE OF METALLIC WIRES. 373 
Desrption of Spsinen | Moke | Gongues | eatge | ar | deme) MD 
tron, No: 1). ole ook i Ske: 
preNor2: 2 .| 4534 64:5 0 812 x 108 242 x 108 0:130 
4 me ‘ ibe ce 34 813 x 10° 248 x 108 0:156 
i. NO: 3.t. . | 4462 88:9 0 OO aes 248, 0°352 
Ca Ae : os in 3:4 Gsoay 253°"; 0-414 
Copper, No. 1, Pl Oe 98:0 0 ASS De i25" 0°136 
s Sik ; bs bad 10°6 489 __, 134, 0:217 
eNO. 2; . | 22°68 92:0 0 427, i257"; 0:332 
German silver, 2). DLS 63:1 0 494 ,, 5a, Oe 0:379 
; 3 : ate Ye 9650) 104980, 159% 5 0-434 
In the appended tables of results the first column gives the date on which 
the experiment was conducted. I give these dates because it seems possible 
that, if the experiment be left for any considerable time, a gradual change may 
take place in the stretched wire, thus rendering the results on different days at 
variance. An examination of the tables shows that in the cases of copper and 
German silver scarcely any change takes place after the first few minutes. In 
the case of iron, however, a very marked change takes place. In consequence 
of this I give time readings for one of the wires extending over about twenty- 
four hours after stretching the wire. These readings are shown in Table V. 
No. 5, and are represented graphically by the curve marked “time variation 
for iron.” This variation is interesting, and will be made the subject of special 
investigation. It is possible to account for the change, either by molecular 
rearrangement in the wire or by oxidation of the surface. I do not think that 
the second can have any sensible influence, as oxidation would increase the 
resistance, both of the standard and of the stretched wire, and since the 
stretched wire had in each case a clean surface, it ought to have been more 
affected than the standard, which was used for several experiments, and had 
been for some time exposed. For these reasons I think that oxidation would 
have had an opposite effect, namely, to show increased instead of diminished 
resistance. The experiment shown in the curve seems to indicate a change 
proportional to logarithm of the time. 
* This high value for the conductivity of this wire was verified by repeated measurements. 
