.s7A' WILLIAM .sY/..]//:.V\, F.R.S. 145 



by atmospheric currents, were prevented by a metallic screen 

 surrounding the flames and the heated box. 



This apparatus is represented in Fig. 1, Plate 12. 



The temperature of the box was gradually raised to 350 Centi- 

 grade, and then lowered ; and observations were taken at regular 

 intervals of increasing and decreasing temperature. 



The results obtained in this further set of experiments are given 

 in the second table. The wire employed was not the same as that 

 employed in the first series, which accounts for certain differences 

 in the ratio of increase observed, although in other respects the 

 accordance of the two series may be considered satisfactory. 



In order to test these discrepancies, a third set of experiments 

 \\as undertaken, with the same platinum wire which had been 

 employed in the second set, with the difference, that the chamber 

 containing the tube and wire and the thermometers was filled 

 with Unseed oil. The results are given in the third table, in two 

 brief series, the object being, in this case, to test the former 

 experiments by a few very careful observations in which the flames 

 were so adjusted by a gas-pressnre regulator, that a perfectly steady 

 heat could be maintained for an hour, or more, to insure identity 

 of temperature in every part of the chamber. 



The general accordance between these results is best shown in 

 the diagram No. 1, Plate 14, where the first, second, and third 

 series of observed results are represented by the lines marked 

 1, i\ and 3, respectively. The horizontal divisions of the sheet 

 represent Centigrade degrees of temperature measured from the 

 absolute zero of temperature ; and the vertical divisions units of 

 resistance divided into tenths. 



With the exception of one observation, which has evidently 

 been taken or noted in error, the accordance between the second 

 and third series of observations is satisfactory. They represent a 

 line, curved downwards towards the X axis, which it crosses at a 

 point near the absolute zero^ or 274 Centigrade below the freezing 

 point of water. 



COPPER, IRON, SILVER, ALUMINIUM. No general conclusion 

 could, however, be drawn from the bearing of one metal. I pro- 

 cured, therefore, wires of comparatively pure copper, of fused iron 

 (or mild steel), of silver, and of aluminium, which were subjected 

 to the same series of observations as before described. The results 



VOL. II. L 



