Specific Electrical Resistance of Copper-Tin Alloys. 555 



The observation of resistance consisted in placing the slider 

 at a definite point on one of the rods, and feeling about on the 

 bridge- wire with another slider for a point of the same potential. 

 The slider on the rod was then shifted to another part of the 

 same rod and the bridge-slider shifted correspondingly. The 

 shifts on the bridge-wire (German silver) were of course rather 

 small ; but they could be read to the tenth of a millimetre ; and, 

 moreover, the resistances in the circuit were so adjusted that 

 any step along the rods corresponded to a step of 3*62 times 

 the resistance along the bridge-wire. It is needless to say 

 that a weak battery and a reflecting galvanometer were used, 

 and that the contacts were only made momentarily. As an ex- 

 ample of the error to be expected I may quote the following. 



Two different steps along rod A gave, as the step on the 

 bridge-wire corresponding to the resistance of 1 centimetre 

 length of the rod A, the numbers 



•0328 and -0325 centim. 

 Two different steps along rod B gave 



•0269 and -0270 centim. 

 Different steps along rod C, which was a trifle less uniform, 

 gave 



•123, -121, -1177, -1173, -1185, '1171 centim. 



As the step *1185 was deduced from the entire length of the 

 rod, I have taken this as the mean. The steps which gaA^e 

 •123 and *121 were short ones; and their probable error is 

 therefore greater than that of the others. Moreover they were 

 made at the thinner end of the not quite cylindrical rod. As 

 regards the mean cross section of the rods, it was estimated by 

 taking a number of readings of the diameters in all parts of 

 the rods by means of a micrometer-gauge reading to the hun- 

 dredth of a millimetre. 



The mean diameter of rod A was '879 centim. 

 v j) ?? B „ -873 „ 



?? 11 11 ^ n *ooJ „ 



The rods had been cast in the same mould ; but C had con- 

 tracted more than the others. 



On the 7th of November Mr. Roberts sent me three more 

 rods, which I labelled D, E, F, and then determined their re- 

 sistance in the same way. 



The mean diameter of these rods, which were perfectly 

 strong and might have been drawn into wire, was 



D -880 centim. 



E -862 „ 



F -872 „ 



These had had their surfaces filed after casting. 



