Metals at High Temperatures. 



385 



measurements upon the silver bar as well as upon all those fol- 

 lowing it were carried out in the second oven tube only. 



Tables lYa and TVb contain the individual measurements. 



Table IV 6. 

 Silver Bar. 



June 22. 



2' (mm.) 



June 23. 



A s (mm.) 

 Observed. Calculated. 



C 

 250 

 500 

 750 



875 



2-352 



2-360 



2-640 



2-640 



2-970 



2-921 





1-649 







2-356 



4-996 



7-941 



9590 







2-356 



5-002 



7-938 



9-515 



The first day's heating produced a permanent increase in the 

 length amounting to 0-027 mm , the second 0'039 mm . The length 

 at the close was 484/l mm at 0°. The formulae are 



A s = 8-844Z + 0-00232£ 2 (m) 

 X = 118270* + 4-793£ 2 |10- 9 



5. Nickel. 



The nickel bar (6 mm in thickness) was at first ruled with five 

 scale divisions on each end like the others and the observations 

 only extended to 750°. In spite of the fact that the metal 

 tarnished when heated, the divisions remained clear and sharp 

 throughout until, at the close of the third series of observations, 

 the temperature was carried up to 1000°. At this temperature 

 not only did the divisions disappear but the surface and the 

 contour lines of the bar could scarcely be distinguished from 

 the oven background. 



To remedy this a small thin platinum plate was prepared, 

 some 5x5 mm in size and 05 mtn thick and shoved into carefully 

 filed dove-tail grooves in the bar so as to sit tightly in place 

 with its polished surface in exactly the same position which 

 the lost marks upon the nickel had occupied. Upon this 

 platinum surface which of course also lay in the axis of the 

 bar, new divisions were ruled and used in the two subsequent 

 series of observations. If we assume from the ratio of the 

 lengths that 1 per cent of the expansion is due to the platinum, 

 the result will be influenced only 0-01(5 : 78-3-57) = 0-02 mm at 

 750°. 



The length of the bar during the first observations was 

 482'6 mm at 0° ; after the insertion of the platinum plates the 

 new sets of divisions proved to be 02 mni further apart. The 

 last observations are corrected for this increase in the length 

 of the bar. 



