[ 442 ] 



L. On the Temperature Coefficient of tlie Modulus of Longi- 

 tudinal Elasticity of Steel. By H. M. Dadourian, Ph.lJ., 

 Associate Professor of Physics, Trinity College, Hartford, 

 Conn* 



TJ^HE effect of temperature upon the modulus of longitu- 

 J~ dinal elasticity of steel has been the subject of many 

 researches. A review of these investigations brings out the 

 fact that the results obtained by different observers differ 

 considerably both in magnitude and in character. 



Kohlrausch and Loomis f, Gray, Blyth and Dunlop J, 

 Harrison §, Dodge ||, and others have shown that the 

 modulus decreases continuously as the temperature is in- 

 creased in the region between 0° C. and several hundred 

 degrees above this temperature. On the other hand, 

 WertheimH, Miss Noyes **, Shakespeare ft? and Walker ft, 

 have observed that as the temperature is increased the 

 modulus increases to a maximum in the neighbourhood of 

 100° C, and then diminishes uniformly. In other words, 

 the temperature coefficient changes sign from positive to 

 negative at about 100° C. Moreover Miss Noyes has found 

 that if sufficiently large loads are used in stretching the 

 wire under investigation, this change of sign disappears and 

 the coefficient becomes negative throughout the temperature 

 region used. Walker's work shows, on the other hand, that if 

 larger loads are used the temperature at which the change of 

 sign takes place is shifted but the effect itself is not eliminated . 

 Both Miss Noyes and Walker were able to observe the 

 change of sign only when the wire w T as heated by passing an 

 electric current through it, and ascribed the effect to this 

 method of heating. Yet Shakespeare was able to obtain 

 the change of sign by heating his specimen by means of a 

 steam jacket. He also found that the effect disappeared 

 after successive heating and cooling for several times. 



A certain degree of divergence in the results is to be ex- 

 pected in view of the fact that the specimens of wire used 

 were all different. But the discrepancies among the values 



* Communicated by the Author. 



t Pogg. Ann. cxli. p. 4il (1870). 



t Proc. Roy. Soc. lxvii. p. 180 (1900). 



§ Phvs. Soc. London, xxvii. p. 8 (1914). 



|| Phvs. Rev. v. p. 373 (1915). 



51 Pogg. Ann. E 2, p. 1 (1848). 

 ** Phvs. Rev. ii. p. 277 (1895) ; iii. p. 432 (1896). 

 tf Phil. Mag. xlvii. p. 539 (1899). 



XX Proc. Rov. Soc. Edin. xxvii. p. 343 (1907) ; xxviii. p. 652 (1908) ; 

 xxxi. p. 186 (1910). 



