Prof. Draper on the Production of Light by Heat. 391 



hot ; and these results might be corroborated by observing the 

 inverse phenomena, when the barrel is taken from the fire and 

 suffered to cool. 



With respect to platinum, brass, antimony, gas-carbon and lead, 

 they all became incandescent at the same time as the iron barrel 

 itself. I could not discover the slightest difference between 

 them, either in heating or cooling ; and it is worthy of remark, 

 that the lead was of course in the liquid condition. But the chalk 

 and marble were visible before the barrel was red-hot, emitting a 

 faint white light ; and the fluor spar still more strikingly so, its 

 light being of a beautiful blue ; and even when the barrel had 

 become bright red I could still see the spar, which had decrepita- 

 ted to a coarse powder, by its faint blue rays. In these cases it 

 was not, however, incandescence but phosphorescence that was 

 taking place. I infer then that all solids, and probably melted 

 nietals, shine at the same thermometric point. 



The temperature of incandescence seems to be a natural fixed 

 point for the thermometer ; and it is very interesting to remark 

 how nearly this point coincides with 1000° of the Fahrenheit 

 thermometer, when Laplace's coefficient for the dilatation of pla- 

 tinum is used. Upon that coefficient the point of incandescence 

 is 1006° F. 



In view of these considerations, and recollecting that the num- 

 ber given by Daniell is 980°, and that of Wedgwood 947°, I be- 

 lieve that 977° is not very far from the true temperature at which 

 solids begin to shine. It is to be understood, of course, that this 

 is in a very dark room. 



I pass now to the second proposition. The rays emitted by 

 the incandescent platinum were next received on a flint glass 

 Prism, placed so as to give the minimum deviation, and after dis- 

 persion viewed in a small telescope. A movement could be given 

 to the telescope, which was read off on an annexed scale. How- 

 ler, instead of bringing the parts of the spectrum under meas- 

 urement to coincide with the wires stretched across the field of 

 the instrument, I found it more satisfactory to determine them by 

 Ringing them to one or other of the edges of the field ; a process 

 ^ r eli adapted to ascertain the position of the extreme rays, the 

 faint light of which contrasted well with the darkness by which 



rt was surrounded. 



spectrum 



points 



accurately 

 s it was 

 that all 



the observations might be brought to a common standard of com- 



parison 



ce nce, such as are in the sunshine and daylight. I therefore pre- 

 viously determined the position of the fixed lines in a spectrum 

 formed by a ray of reflected daylight which p.- sed throt h a 

 fissure A of an inch wide and one inch long, occupying exactly 



