Properties, Specific Resistance, and Hardness of Steel. 365 

 Table IX. 



Remarks. 



Volume. 



Increase of 

 same. 



T. E. H. 



Soft 



1-000 

 1-002 

 1-005 

 1-010 



o-ooo 



0-002 

 0-005 



o-oio 



0-000000 

 0-000024 

 0-000042 

 0-000117 



"Blue" 



"Yellow" 



"G-l ass-hard" 





In the fourth column the T. E. H. of rods cited in Table III. 

 (these, as I believe, corresponding very nearly to those for 

 which the data of Froinme apply) is added. If we take into 

 consideration that the results were obtained from different 

 material, the parallelism observable is striking. To the very 

 large difference between glass-hard and yellow-annealed, 

 when compared with the much smaller difference between 

 yellow and blue, blue and soft, as seen in both observations, I 

 would once more call attention. The fact that glass-hard rods 

 can be considerably annealed at comparatively Ioav tempera- 

 tures, must be regarded as an adequate indication of their un- 

 natural strained condition. 



A second result of Froinme, that the specific gravity of thin 

 rods suffers greater loss by glasshardening than that of thick 

 rods, also harmonizes with the conclusions drawn in § VII., b, 

 with reference to the T. E. H. of such bars. 



VIII. Hardness and Specific Resistance of Steel. 



From the data found for specific resistance of steel rods 

 of different hardness, inferences analogous to the above may 

 be deduced : — 



a. The specific resistance of steel increases continuously with 

 its mechanical hardness. 



b. Rods like-annealed differ but slightly, glass-hard rods 

 considerably with respect to their specific resistance. 



AS 



c. On comparing the values found for the ratio m F Vp 



we infer that the specific resistance of steel is approximately a 

 linear function of its thermoelctric hardness. In fig. 5 these 

 results are graphically represented. 



I will remark, however, that the assumption of propor- 

 tionality as based on the above figures is to be regarded as a 

 first approximation only, notwithstanding the fact that the 

 discrepancies fall within the errors of experiment. A rigid 

 discussion of the latter comes more appropriately within the 

 scope of another paper, soon to appear. In this place I would 



