483 



Prof. C. Barus on the Secular Softening 



degrees centigrade relative to silver, and % the specific resist- 

 ance of steel (microhms, cubic centim. 0° (J.), so that 15*176 

 would be the thermo-electric power of a steel rod whose resist- 

 ance is zero. 



Table I. — Change of Specific Resistance of Grlass-hard 

 Stubbs' Steel Rods with the Temperature and Time of 

 Tempering. 



Temperature. 

 °0. 



Time. 

 Minutes. 



s . 



Microhm, 



0°O. 



Temperature. 



°C. 



Time. 

 Minutes. 



Microhm, 

 0°C. 



60 







408 



185 







39-1 





60 



394 





10 



26-7 





1-20 



388 





30 



25-4 





180 



38-3 





60 

 120 



24-7 

 243 



100 





 10 



39-0 

 36-4 





180 



23-7 





30 



345 



330 







377 





60 



33-2 





1 



172 





120 



31-8 





30 



16-8 





180 



31-1 



1000 



90 



16-7 

 150 



Thus in Table I., s is the specific resistance of glass-hard 

 steel, exposed to the temperature stated in the first column, 

 continually during the time t in minutes given in the second 

 column. The value s is reduced to zero centigrade. 



The results are graphically given in the accompanying 

 chart, where the abscissas are minutes of time of exposure, 

 and the ordinates the corresponding resistances s , when the 

 rod is tempered at the temperature marked on the curve. 



Clearly these curves (all exponential in character) are the 

 expression of a single law giving the relation of resistance to 

 time and temperature ; but with the investigation of this func- 

 tion I am not further concerned here. The features to be 

 noted are the occurrence of a single and definite degree of 

 soft temper for each temperature. This limit of softness is 

 approached more rapidly in proportion as the temperature to 

 which the hard rod is continually exposed is higher. 



4. In correlative experiments Prof. Strouhal and I further 

 showed that a hard rod which has been tempered to the limit 

 of softness for any temperature, say the boiling-point of 

 water, is then molecularly stable at a lower temperature, and 

 this in a more marked degree in proportion as the latter 

 temperature falls below the temperature at which the rod is 



