412 CARNEGIE INSTITUTION OF WASHINGTON. 



600° and 1000° C. The effect is also present in many sulphides, par- 

 ticularly sidot blende (ZnS), Balmain's paint (CaS), and some of the 

 phosphorescent sulphides of Lenard and Klatt. It is likewise found 

 in various other compounds, such as calcium carbonate, cadmium 

 phosphate, boric acid, telluric acid, and in crystals of calcite, fluorite, 

 kunzite, synthetic ruby, etc. 



Many preparations that are highly fluorescent at ordinary tempera- 

 tures, such as calcium tungstate, artificial willemite, canary glass, 

 didymium glass, etc., fail to respond to excitation at these high tem- 

 peratures, and the oxides of iron, copper, cadmium, titanium, nickel, 

 cobalt, barium, etc., are altogether inactive. 



A New Source of Excitation. 



This luminescence at high temperatures is not photo-luminescence — 

 that is, it can not be produced by exposing the material, heated to the 

 required temperature, either to the iron spark, the radiation from an 

 electric arc, or to sunlight. The oxides mentioned above are not 

 excited by light at any temperature, and the sulphides, which are 

 powerfully photo-luminescent with persistent phosphorescence when 

 cold, cease to be excited by light at much lower temperatures, as was 

 abundantly established by Lenard and Klatt in 1904. 



The glow is not merely a type of selective temperature radiation. 

 Heating alone will not produce it. Excitation at these temperatures is 

 obtained by contact with the hydrogen flame and we have called this 

 new source of excitation flame excitation. It occurs only in a certain 

 active zone within the flame along the boundary between reduction 

 and oxidation. The powerful reducing action of the hydrogen flame 

 seems to be essential, all other flames thus far tried being ineffective. 



A full supply of free oxygen in the atmosphere is also necessary, it 

 being possible to check the glow by a surrounding mantle of the fumes 

 from a Bunsen burner or to enhance it by a gentle current of cold air. 



The close relation of this effect to luminescence at ordinary tempera- 

 tures is established by the following observations: 



(1) The fluorescence is followed, after extinction of the flame, by 

 phosphorescence which, in the only cases thus far studied, is of the 

 vanishing type with the usual linear processes; total duration a few 

 hundredths of a second. 



(2) The spectrum does not correspond in distribution of intensities 

 w^ith that of temperature radiation, but consists of two or more broad, 

 overlapping bands. These, as shown by detailed spectro-photometric 

 measurements made upon calcium oxide and upon certain of the sul- 

 phides by Dr. H. L. Howes, are made up of numerous equidistant com- 

 ponents identical with those belonging to the ordinary luminescence 

 spectrum of the substance in question. 



