68 



KENNELLY-SANBORN— INFLUENCE OF 



[April 24, 



O.I 14 mm. (0.0045 inch). In the tests of 1909, copper wires were 

 used. The advantage of copper is that its resistivity temperature- 

 coefficient is relatively large, and is fairly reliable. On the other 

 hand, hot copper wires oxidize superficially when driven through 

 the air, and are therefore subject to variation in convective dissipa- 

 tion, owing to this change of surface condition. As the test-wire in 

 the new measurements had to be driven inside a steel tank, with 

 only occasional inspections, it was decided to employ platinum, in- 

 stead of copper ; although the resistivity temperature-coefficient of the 

 platinum was but little more than half that of copper; so that the 

 resistance of such a platinum wire is not so sensitive to changes of 

 temperature as a copper wire. Consequently, greater care was 

 needed in the electrical measurements of resistance in the platinum 

 test-wire, in order to determine the temperature elevation. 



A measurement of the temperature-coefficient of resistivity of 

 the platinum wire used was made by immersing 5.5 meters of it on 

 a reel in an oil-bath, and measuring the resistance at twelve different 

 temperatures between 0° C. and 100° C. As shown in Fig. 6, the 

 results obtained lie close to the straight line : 



pj = Po(i+o.oo25750 



absohm-cm ( i ) 



where pt is the resistivity at ^° C. (absohm-cm.), and p^ is the 

 resistivity at 0° C. (absohm-cm.). 



The particulars concerning the test wire are given in the accom- 

 panying Table: 



TABLE L 

 Test-Wire Dimensions and Data. 



Test-Wire Holder. 

 The test-wire was held in a fork or frame, mounted on the shaft 

 of the driving motor. The fork is indicated in Fig. i. It is counter- 

 poised by the sliding weight /. The test-wire is shown at b, held 



