168 



SCIENCE 



[N. S. Vol. XXXI. No. 788 



If such salts as calcium fluoride were 

 conductors at ordinary temperature, useful 

 electrodes for flame arcs would probably be 

 made from them. Such conducting ma- 

 terials as iron oxide, carbides, etc., have 

 been used for flame arc electrodes, and a 

 great many of the so-called magnetite arcs 

 are now in use. The electrodes in this case 

 are largely magnetic oxide of iron, with 

 such other ingredients as titanium and 

 chromium oxides, to increase the intensity 

 of light, to raise the melting-point of the 

 mixture, etc. 



As will be seen from observing this arc, 

 the light is very white and intense and is 

 generated by the heated vapors of the arc 

 proper. A great many modifications of 

 this arc principle are possible. Titanium 

 carbide and similar substances give charac- 

 teristic arcs, and some of them are very 

 intense and efficient. For purposes of com- 

 parison, I have added to this illustrating 

 experiment an arc of titanium carbide and 

 one of copper. 



THE NEENST LAMP 



A distinct species of electric incandescent 

 lamp is that invented about ten years ago 

 by the well-lmown physical chemist. Pro- 

 fessor Nernst. This employs for filaments 

 a class of bodies which are not electrical 

 conductors at all at ordinary temperatures, 

 and which, at their burning temperatures, 

 do not conduct the current as metals and 

 carbon, but as a solution does. This kind 

 of conductivity, the electrolytic, involves 

 electrochemical decomposition at the elec- 

 trodes, and in the case of the Nernst fila- 

 ments these otherwise destructive reactions 

 are rendered harmless by the continual 

 oxidizing action of the air. For this reason 

 this type of lamp will not burn in vacuo. 

 For its most perfect utility the principle of 

 the Nernst lamp seems to require a mixture 

 of oxides, because a single one is not so 



good a conductor nor so luminous. It uses 

 oxides because these are the most stable 

 compounds known, and it uses the rare 

 earth oxides because they have higher melt- 

 ing-point than other oxides. As the effi- 

 ciency very rapidly rises with temperature, 

 there is a great advantage in using the most 

 infusible base possible. For that reason, 

 zirconia, thoria, etc., are usually employed. 

 In this lamp a rod or filament of an oxide 

 mixture, much like those used in Welsbach 

 mantles, is heated by the current externally 

 applied until it reaches a temperature at 

 which it becomes a good conductor itself. 

 Here again the peculiar laws of light radia- 

 tion are illustrated, the light emitted at a 

 given temperature being determined by the 

 nature of the substance. Just as the pure 

 thoria gives a poor light compared to the 

 mixture with one per cent, ceria, so a pure 

 zirconia rod, heated by the current, gives 

 much less light than a rod containing a 

 little thoria, ceria or similar oxide. Work 

 done by Coblentz on the energy-emission of 

 such rods shows the emission spectra, at 

 least in the infra-red, to vary with the 

 nature of the substance. In general, the 

 spectra are not continuous like the spectra 

 of metals and black bodies, but seem to 

 occupy an intermediate position between 

 these and luminous gases, which we know 

 have usually distinct line spectra. 



This recalls the subject of selective emis- 

 sion. Coblentz has shown selective emis- 

 sion in the long wave-lengths for a Nernst 

 glower. This is shown in comparison with 

 the emission of a black body, in curve 

 No. 1. The two sources, when compared 

 at the temperatures where they exhibit the 

 same wave-length for maximum emission, 

 differ very considerably in emission in the 

 infra-red, the black body giving more en- 

 ergy at the blue end and less at the red 

 end of the spectrum. 



This is still more noticeable in the curves 



