362 Dr. C. Bams on the Relation between the Thermoelectric 



that the change of state due to glasshardening is chemical in 

 its nature. Li this opinion I believe most chemists at pre- 

 sent also concur, it being assumed that the uncombined carbon 

 in soft steel is, by sudden immersion, converted into the che- 

 mically combined. In summing up the facts by which the 

 hypothesis is furthermore supported, we will mention, in the 

 first place, the detailed analogy* which exists between the 

 white pig used in the manufacture of Bessemer steel and glass- 

 hard steel, on the one hand, and ordinary (grey) cast iron and 

 soft steel on the other, the former containing carbon only in 

 the combined, the latter also in considerable proportion in the 

 uncombined state. Secondly, hardening by a process of wire- 

 pulling, hammering, &c, will in all probability increase the 

 specific gravity of steel ; hardening by sudden cooling, as is 

 well known, diminishes it. In the former case, the thermo- 

 electric current usually passes from soft to hard through 

 warmf ; in the latter, in a contrary direction. Indrawn wire 

 the specific resistance is smaller^ in hard-tempered, and greater 

 than that of the same wire in the soft state. 



Considering these facts as a whole, we are perhaps justified 

 in distinguishing between a process of chemical and a process 

 of mechanical hardening. This, however, does not prevent us 

 from paying due regard to the series of physical phenomena 

 which always accompany the former. To these the peculiar 

 internal structure of glass-hard bars, the warping which fre- 

 quently attends sudden cooling &c, are to be referred. "We 

 conclude, therefore, that the cause chiefly influential in bringing 

 about glasshardness in steel is chemical in its nature, and that 

 in consequence of the physical phenomena which invariably 

 accompany it the degree of glasshardness is more or less mo- 

 dified. On the latter ground the continual increase of the 

 T. E. H. after the critical temperature above referred to has 

 been reached is to be explained. 



d. The observed T. E. H. can of course only be assumed as 

 directly expressive of its hardness when the rod under experi- 

 mentation is homogeneous throughout. This is approximately 

 the case for thin rods. In thick bars which in the glass-hard 

 state may be considered as made up of a series of concentric 

 cylindrical shells, the hardness of which decreases rapidly as 



* See Wagner's Chem. Technol. 8 ed. pp. 14, 15, 29. These analogies 

 refer to mechanical properties, method of preparation, colour. 



t Magnus, Pogg. Ann. lxxxiii. p. 469, 1851 j Sir W. Thomson, Phil. 

 Trans, iii. p. 722, 1856. 



| Mousson, N. Denkschr. tier schweiz. Gesellsch. xiv. 8. pp. 1-90. The 

 reader is also referred to Chwolson, Carls. Repert. xiv. p. 15. 



