ELECTRICAL RESISTANCE UNDER PRESSURE. 



79 



description of the capillary and the method of filling it will be found 

 under potassium. Seven runs were made with the sodium in glass, 

 with several different capillaries. These runs were at 134.0°, on both 

 solid and liquid, at 143.6° on the liquid only, at 163.1° on the liquid 

 only, at 165.0° on both liquid and solid, and at 171.8°, 197.1° and 198.1° 

 on the liquid only. In addition, readings were made at atmospheric 

 pressure from which the change of resistance on melting and the 

 temperature coefficient of resistance of the liquid could be obtained. 

 These runs were of varying degrees of excellence. Those at 165° and 

 197° showed zero shifts of only 0.5% and 0.3%, and were given the 

 greatest weight in computing the resistance of the solid. In general 

 the observed points lay very smoothly, and there was little diiference 

 between the results with increasing and decreasing pressure, indicating 

 little direct effect due to the constraint exerted by the walls of the 

 capillary. 



100° 



1 20° 1 40° f60° 

 Temperature 

 Sodium 



80° 



Figure .3. The ratio of the resistance of Hquid to solid sodium as 

 function of temperature along the melting curve. 



The ratio of resistance of liquid to solid was observed at three 

 temperatures; the observed points are shown in Figure 3. The ratio 

 decreases somewhat with increasing pressure and temperature along 

 the melting curve. In reducing the observed values to smoothness, 

 it was assumed that the ratio varies linearly with temperature, and is 

 given by the line shown in the figure. 



The values of resistance as a function of pressure and temperature 

 from 0° to 200° and to 12000 kg. are shown in Table III. This table 

 falls into two parts; the first part, including the values through 100°, 

 are relative values of the "observed" resistance, being derived from 

 measurements on the bare wire. Above 100° the values listed are 

 relative values of the specific resistance, the observations ha\ang been 

 made on the sodium in a glass capillary, and corrections applied for 

 the compressibility and thermal expansion of the glass. For the 



