588 



BRIDGMAN. 



the maximum deviation of any point from the smoothed curves is 

 0.4% of the pressure effect, and the average numerical deviation is 

 0.07%, or, excluding the run at 0°, where the effect of viscosity was 



c 006 



o 



la 



t .005 

 Q 



I .004 

 I 

 ^ .003, 



-i .0>95 



Pressure, Kg./Cm.' X 10' 



Figure 4. Tin, results for the measured resistance. The deviations from 

 linearity are given as fractions of the resistance at kg. and 0° C. The pres- 

 sure coefficient is the average coefficient between and 12000 kg. 



apparent, 0.03%. The deviation curves are sensibly not symmetrical, 

 although the pressure of maximum deviation is very nearly the mean 

 pressure. 



The average temperature coefficient of tin between 0° and 100° 

 was 0.00447 This is not quite as high as the value of Jaeger and 

 Diesselhorst ^ for tin from the same source as this, which by their 

 analysis had less than 0.03% lead. Their value for tin rod was 

 0.00459 (corrected for range as explained in the introduction). 



The initial value of the pressure coefficient at 0° has been found by 

 Beckman ^ to be — O.O592. This is to be compared with -O.O4IO4 

 given above. The average coefficient to 12000 kg. agrees very closely 

 with Beckman's value. 



The general character of the results is the same as for indium. If 

 the curves of resistance against pressure are so changed in scale that 

 the resistance at kg. for each temperature is the same, then the 

 curves for the higher temperatures are the steeper and have the greater 

 curvature. 



Thallium. This was electrolytically prepared from two samples 

 of metal which originallv came from Merck and Eimer and Amend, 



