of Cold Hard Steel. 487 



By way of introduction I will therefore compute a table of 

 the resistances of tempered steel from data obtained by Prof. 

 Strouhal and myself seventeen years ago *, as it will aid mate- 

 rially in the interpretation of the present results. 



2. Steel is a well chosen metal for the purpose, since its 

 electrical resistance may by the process of tempering from 

 soft to hard be made to increase more than threefold. One 

 is thus able to trace the secular variations on this enor- 

 mous scale with certainty, and without reference to concomi- 

 tant secular changes of the resistance of the german-silver 

 standards with which the steel rods are compared. The 

 resistance of these standards cannot have varied more than a 

 fraction of 1 per cent., whereas the steel rods show a change 

 of about 15 per cent, during the interval of observation. 

 Of course the same standards and the same steel rods were 

 used throughout. For details as to their mass constants, 

 homogeneity, methods, &c, the reader may consult the 

 previous paper (I. c). 



The effect of a drawn strain imparted to a metal on its 

 electrical resistance is fortunately of very much smaller order 

 as compared with temper in steel. Nevertheless where elec- 

 trical standards are in question these small changes are a 

 serious discrepancy. Much of this can be removed by judicious 

 treatment I dare say, since I showed that the mechanical 

 strain or any molecular change varies in the lapse of time 

 and with temperature under like conditions with temper in 

 steel f. 



3. The following Table I. contains the results for the change 

 of the specific resistance, s Q , of originally glass-hard Stubbs* 

 steel rods with time and temperature. These data are the 

 means of closely accordant values of three rods for each tem- 

 perature. Instead of reducing the values for resistance directly, 

 the object in view was more easily attained by using the 

 thermo-electric chart which Prof. Strouhal and I constructed, 

 inasmuch as we showed that the thermo-electric power of steel 

 is subject to similarly phenomenal variations with temper ; 

 and that the relation of the thermo-electric power of steel 

 referred to silver and the specific electrical resistance of steel 

 (over ninety rods were examined) is expressed by the equation 



y= 15-176 -0'4123;& 



where y denotes thermo-electric power in microvolts and 



* Wied. Ann. xi. p. 930 (1880); Bull. U. S. Geolog. Survey, No. 14 

 p. 55 (1885). 



t This Magazine, Feb. 1889, p. 155 ; Bulletin U. S. Geolog. Survey, 

 No. 94, p. 17 etseq. (1892). 



