250 Mr. H. Tomlinson on some 



of annealed iron is temporarily decreased to the extent of 2*693 

 per cent., when the temperature is raised from 0° to 100° C* 



Again, an annealed pianoforte-steel wire, when tested with 

 a certain load at the temperature of 12° C, was temporarily 

 elongated to the extent of 1*502 half-millimetres; when heated 

 to 100° C. in an air-chamber the elongation was 1*487 half-milli- 

 metres; and when cooled again and tested 24 hours afterwards, 

 was elongated by 1*450 half-millimetres f. Thus, as with the 

 torsional elasticity, there was a permanent increase of elasticity , 

 and a temporary decrease of 2*58 per cent. Time in this case 

 also is an important element, for the elasticity immediately 

 after cooling was very appreciably less than when a long rest 

 had been given. 



Similarly, I have shown both for longitudinal and torsional 

 elasticity that an iron wire, after having been permanently ex- 

 tended by traction, has its elasticity very perceptibly increased 

 by long rest. It is also well known that the portative power of a 

 magnet can be considerably increased by putting on the load by 

 small quantities at a time, with long intervals of rest between. 



These and other considerations prove beyond a doubt, that if 

 the molecular arrangement of iron be disturbed by any kind of 

 stress whatever, exceeding a certain small limit, the molecules 

 will not assume at ordinary temperatures those positions which 

 will secure a maximum of elasticity, until a rest of many hours 

 has been given. 



The Velocity of Sound in Iron. 

 According to Wertheim, the velocity of sound in iron and 

 steel is increased by a rise of temperature not extending beyond 

 100° G.J Now in no sense whatever is this statement correct. 

 It is true that the longitudinal elasticity of iron, as determined 

 by the method of statical extension, may be found greater at 

 100° C. than at 0° C, provided w T e begin with the lower tempe- 

 rature first and the wire has not been previously tested at 

 100° C. But, as we have seen, the apparent temporary 

 increase of elasticity is really a permanent one ; and if the wire 

 be repeatedly heated to 100° C. and then cooled, subsequent 

 tests will always show a less elasticity at the higher tempe- 

 rature than at the lower, if sufficient rest after cooling be 

 allowed. When, however, we come to such small molecular 

 displacements as are involved in the passage of sound through 

 a wire, even the apparent increase of elasticity, mentioned 

 above as taking place at the first heating, vanishes. I have 

 been able to prove that, when an iron or steel wire is thrown 



* Proc. R. S. No. 244, 1886. 



t Phil. Trans. Part I., 1883, p. 130. 



X Ann. de Chimie et de Phys. 3me 8 6rie, 1844, p. 421. 



