186 C. Barns — Effect of Magnetization, etc. 



Table 6. 

 Rigidity of magnetized Stills' 1 steel. w = 0-0043. 



Temper. 



<p x 10 lj 



fix 10 3 



Glass-hard 



3 



4 

 4 

 22 

 25 

 41 

 95 



0-06 



Annealed 100° . .. 



0-08 



190° 



360° 



0-08 

 0-42 



450° 



Soft steel 



0-48 



0-78 



Soft iron . 



3-63 







tance during the second phase.* In other words, if we suppose 

 the wire to pass continuously from hard to soft, the increase of 

 the magnetic coefficient of rigidity, p., is particularly pronounced 

 after the variations of the thermo-electric, the galvanic and the vis- 

 cous properties of steel have practically subsided, jj. therefore 

 shows close affinities to the induced and to the residual mag- 

 netization of an iron-carburet. Agreeing with Wiedemann's 

 results, fx is invariably smaller for steel than for iron. In gen- 

 eral if p. be regarded in its dependence both on co and on hard- 

 ness, it appears that the increment of rigidity of an iron-carbu- 

 ret produced by magnetization is greater in proportion as the 

 metal shows greater tendency to assume permanent set — a 

 result which applies for iron and for steel. 

 Lab. U. S. G. S., Washington, May, 1887. 



* In a letter to the Journal (xxxiii, p. 308), Prof. W. F. Barrett has taken excep- 

 tion to certain remarks made by Dr. Strouhal and myself (this Journal, xxxiii, p. 

 35) on phenomena more or less connected with the second phase of annealing. 

 We regret exceedingly to have overlooked Prof. Barrett's papers. Our object, 

 however, was only to give an enumeration of such observations as had occurred 

 to us incidentally. Of Gore's discovery we were aware, and the statement is thus 

 made in the text and in TJ. S. G. S., Bull. 14. Owing to a misconstruction of the 

 foot-note in the Journal (1. c, p. 35, 1), "ibid." happens to refer to " Wied. Ann." 

 instead of to " Phil. Mag." The reference is otherwise correct. 



We avail ourselves of the present opportunity to state that in Prof. J. A. 

 Ewing's paper on the " effects of stress and magnetization on the thermo-electric 

 quality of iron" (Phil. Trans., II, p. 361, 1886) mention is made of our results only 

 in a final note dated Sept. 17, 1886. Our work was accessibly published much 

 earlier (Wied. Ann., xiv, p. 54, 1881 ; cf. also ibid., vii, p. 408, 1879). We believe 

 we were the first to actually measure Thomson's thermo-electric effect of 

 magnetization as weU as to point out its probable relations to the strain of a mag- 

 netized rod. 



