104 Mr. Herbert Tomlinson on 



strain resulting from thermal agency, and is the strain with 

 which we have now to deal. Such a strain is not peculiar to 

 glass ; for the author has found that a coil of zinc wire, which 

 had been previously heated to nearly the melting-point of the 

 metal and afterwards suddenly cooled by plunging into cold 

 water, continued to show decrease of volume for some minutes 

 after the cooling. The internal friction of a metal seems to 

 be intimately connected with the subpermanent retentivity ; 

 and the author has already investigated the internal fric- 

 tion of several metals at various temperatures*. The results 

 of the above-mentioned investigation show that changes of 

 temperature, even when they do not exceed 100° C, exercise 

 a very considerable effect on the internal friction ; and it 

 seemed likely that a preliminary examination of the internal 

 friction of iron at very high temperatures might be of service 

 in elucidating the problem of recalescence. 



The Internal Friction of Iron at High Temperatures. 



Experiment I. — In fig. 1, a b is an iron wire 49*5 centim. 

 in length and '041 centim. in diameter ; the ends a and b are 

 brazed to two pieces of rather stout brass wire ; the upper 

 piece of brass wire is furnished with a binding-screw, S, and 

 the lower one is soldered to the centre of a hollow horizontal 

 brass bar, V ; D is a piece of copper wire soldered at its 

 upper extremity to V, and dipping with its lower extremity 

 into mercury contained in a glass vessel, H. The wire hangs 

 in the axis of a glass tube, T, fitting into the cover of a box, 

 B ; the tube is closed with a cork, C, and the box B, though 

 provided with a door which can be opened for making adjust- 

 ments or for starting the torsional oscillations, is kept shut 

 up during the actual observations. The box is also provided 

 with a glass window, which enables an illuminated circle 

 crossed by a fine vertical dark line to be thrown on to the 

 light mirror, M, and hence be reflected back on to a scale 

 according to the usual arrangement. The wire was set in 

 torsional oscillation by means of a light feather gently pressed 

 against one end of V and then removed f ; the amplitudes of 

 the vibrations were never allowed to exceed one degree. 



The wire was heated by means of an electric current from 

 a battery of 30 Grove cells, an ampere-meter and a set of 

 resistances serving respectively to measure and to vary the 



* Phil. Trans, vol. clxxvii. (part ii. I860) p. 801. Proc. Roy. Soc. 

 No. 244 (1886), p. 343. 



t It would have been better to have had a small piece of iron attached to 

 V and to have started the vibrations by a magnet on the outside of the box ; 

 since the amplitudes diminished very rapidly at the higher temperature. 



