20 Messrs. Nichols and Snow on the Character of 



reflected by the oxide of zinc (p. 420). The results obtained in 

 that investigation led us to the conclusion that certain fluctua- 

 tions in reflecting-power were indications of peculiarities in 

 the law of the effect of heat on the radiating-power of the 

 oxide. In the present paper we purpose to describe an attempt 

 to verify that conclusion by the application of the same method 

 of observation to the study of the visible radiation at tem- 

 peratures between 650° and 1000° C. 



Almost the only material, the radiation of which as a 

 function of the temperature has been determined quantitatively, 

 is platinum. In 1879 one of the present writers described some 

 photometric measurements of the light emitted by that metal*. 

 In the same year Violle f published data concerning the total 

 radiating-power, between 800° and 1775° C, without entering 

 upon the question of the changes in the quality of the light. In 

 the same year also Jacques J attacked the problem of the dis- 

 tribution of energy in the spectrum of various incandescent 

 solids, extending the investigation to platinum and certain 

 oxides of copper, iron, chromium, and aluminium. Dr. Jacques 

 used the thermo-pile and galvanometer in his measurements, 

 and dealt more particularly with the energy-curves of the 

 infra-red. In 1881 Violle|§ published readings for four regions 

 of the visible spectrum (*656, *589, '535, "482), the source of 

 light being platinum at 775°, 954°, 1045°, 1500°, and 1775°. 

 The source of comparison was the car eel lamp. 



None of the existing data gave the law for platinum with 

 sufficient definiteness for our purpose, winch was to compare 

 the radiation of the zinc oxide, wave-length by wave-length, 

 with that emanating from platinum at the same temperature, 

 throughout the entire range of temperatures already indicated. 

 Measurements of the platinum spectrum were according! v 

 made at several stages of incandescence. From these measure- 

 ments a set of curves were plotted, which gave graphically 

 the increase of radiating-power with rise of temperature 

 for each of nine selected regions of the spectrum. The ordi- 

 nates of these curves are relative light-intensities, the abscissae 

 are degrees of the centigrade thermometer. For convenience, 

 the intensity of the region of the D line of Fraunhofer, when 

 the metal is at the temperature of 1000°, has been selected as 

 unity. The comparison lamp was an incandescent lamp held 



* E. JL. Nichols, Ueber das von gluhendem Platin ausgestrahlte Zickt. 

 Dissertation : Gottingen, 1879. 



+ J. Violle, Comptes Bendus, lxxxviii. pp. 171-17-3. 



t W. W. Jacques, "Distribution of Heat in the Spectra of various 

 Sources of Radiation. ' ? Proc. Arner. Acad, of Arts and Sc, 1879. 



§ J Violle, Co mptes Rend us, xcii. pp. 866 and 1206. 



