MEMOIR I.] THE RADIATIONS OF IGNITED BODIES. 25 



example, platinum at 2600 emits almost forty times as 

 much light as it does at 1900. 



The source of light I have employed is in all in- 

 stances a very thin strip of platinum, 1.35 inch long, 

 and .05 of an inch wide, brought to the temperature 

 under investigation by a voltaic current. Platinum was 

 selected from its indisposition to oxidize, and its power 

 of resisting a high temperature without fusion. 



The strip of platinum thus to be brought to different 

 temperatures by an electric current of the proper force 

 was fastened at one end to an inflexible support, and 

 at the other was connected with a delicate lever- index, 

 which enabled me to determine its expansion, and there- 

 by its temperature. For this purpose I have used the 

 coefficient of dilatation of Dulong and Petit. The tem- 

 peratures here given are upon the hypothesis of the in- 

 variability of that coefficient at all thermometric degrees; 

 they are therefore to some extent in error. 



In Fig. 1, a b represents the strip of platinum, the 

 upper end of which is soldered to 

 a stout and short copper pin, a, 

 firmly sunk in a block of wood, c, 

 which is immovably fastened to the 

 basis, d d, of the instrument. A 

 cavity, 0, half 

 an inch in di- 

 ameter is sunk 

 in the block <?, 

 and into this 

 cavity the pin a projects, so that when the cavity is filled 

 with mercury a voltaic current may be passed through 

 the pin and down the platinum. The other extremity 

 of the platinum, , is fastened to a delicate lever, b f, 

 which plays on an axis at g, the axis working in brass 

 holes supported on a block, li. Immediately beneath 



Fig. 1. 



