Expansion and Contraction in Wires under Stress. 315 



are hung side by side from the same top support ; the support 

 being a brass plate with small holes drilled in it through which 

 the wires are drawn and into which they are soldered. The 

 brass plate just mentioned closes the top of the tube loosely ; 

 and it is closed at the bottom also, except for small apertures 

 through which the wires pass perfectly freely. A plentiful 

 supply of steam from the University heating-apparatus enables 

 me to raise the temperature of the whole tube in two or three 

 minutes to the condensing-point of steam ; and a tube (shown 

 in the figure) carries away the condensed water. 



The wires which I have used up to the present time have 

 been fine copper wires,, about No. 34, *008 inch in diameter 

 (similar in size to the wires used in the secular experiments in 

 the tower) ; and, as in secular wires, one of them carries about 

 half its breaking- weight, the other something less than one 

 tenth. The weights are rectangles of thin lead ; and each has 

 little feet of brass wire which rest very lightly against oppo- 

 site sides of a vertical glass plate (marked pp in fig. 1). This 

 glass plate prevents the lead rectangles from turning round or 

 getting into torsional oscillation. 



Before describing the measuring-apparatus, it will be con- 

 venient for me to say a few words as to the requirements, 

 and as to a difficulty which is met with in carrying out the 

 experiments. The wires having been suspended as described 

 above and their weights attached to the lower extremities, a 

 scale divided into half-millimetres was attached to one of 

 them (the lighter), and an index or pointer moving over the 

 scale was attached to the other. To aid in reading, a lens 

 or Quincke microscope-kathetometer (shown in the figure) 

 could be used, though at the commencement this was un- 

 necessary. Thermometers were also inserted into holes in the 

 tube at different heights. 



As soon as the steam was allowed to enter the tube both 

 wires were seen to extend, and the heavily weighted wire ex- 

 tended much more than the lightly weighted wire. The steam 

 was turned off after about ten minutes, and the tube and wires 

 allowed to cool. In an hour or more, when it was quite cer- 

 tain that everything had returned to the initial temperature, 

 the marks were examined again, and it was found that the 

 wires had not returned to their original length ; but had suf- 

 fered a considerable amount of permanent elongation, the 

 heavily weighted wire having suffered more elongation than 

 the other. This was to be expected ; but I. was not prepared 

 to find that, when the heating was repeated again and again, 

 the permanent elongation steadily increased, though each time 

 by a perceptibly diminishing amount. 



As soon as this was found I determined to add temporarilv 



Y2 ^ 



