Resistance of Metals and Alloys. 275 



and at any two points erect perpendiculars to it which are 

 proportional to the resistance of the platinum wire at 0° C. 

 and 100° C, and draw through the summits of these ordinates 

 a straight line cutting the axis of temperature, we may define 

 temperature as follows : — Divide the horizontal interval be- 

 tween the ordinates into one hundred equal parts, and continue 

 the division on the scale backwards to the point at which the 

 inclined line cuts the temperature-axis. Measure the resist- 

 ance of the platinum at any temperature and find the ordinate 

 representing this resistance : the foot of this ordinate marks the 

 temperature on the platinum-scale. Callendar has shown that 

 the difference between temperatures measured on this platinum- 

 scale and the corresponding air-thermometer temperature is 

 very nearly represented by a parabolic function of the air-ther- 

 mometer temperature. It remains to be seen if all pure platinum 

 wires give the same platinum temperature. We have found 

 slight differences between the specific resistance and tempe- 

 rature-coefficients of platinum wires of different diameters, 

 prepared with the greatest care and supposed to be equally 

 pure, but which differences do no doubt depend upon slightly 

 different degrees of hardness or purity. These differences 

 are sufficient to make small but quite evident differences 

 between the platinum temperatures as determined by these 

 different platinum wires. These differences, however, are not 

 more marked than would be the case with temperatures mea- 

 sured by different mercury-in-glass thermometers. We have 

 accordingly selected one of our platinum wires as a standard 

 thermometer and have referred all our measurements of tem- 

 perature to it, and subsequently we may determine the 

 equivalent temperatures on the air-thermometer scale. This 

 platinum wire w r e refer to subsequently as the " thick pla- 

 tinum " wire. In the chart appended to this paper the 

 abscissae of temperature are therefore platinum-temperatures 

 as measured by our standard platinum wire. The general 

 nature of the measurements made consisted, therefore, first, 

 in the measurement of the length and mean diameter of a 

 fine wire of a metal or alloy of carefully prepared qualitv ; 

 secondly, in the measurement of the resistance of this wire 

 when immersed in a bath giving it the necessary temperature; 

 and thirdly, the determination of this temperature of the bath 

 by the measurement of the resistance of the standard platinum 

 wire before and after the measurement of the given wire. 

 From the results of these observations we could deduce the 

 volume-specific resistance of the material at certain platinum 

 temperatures, and plot a series of curves to represent the 

 variation of this resistance with temperature. 



