1080 



lower temperatures. At the high temperature at which the filament 

 glows, practically no reaction takes place as according to Langmuik 

 the intermediate products necessary for tlte formation of WCl^ are 

 dissociated again (under formation of 67 atoms) before the reaction 

 by which WCl, is formed has been completed. 



There are however always parts of the filament and often also of 

 the tungsten supports that carry the wire, at a temperature, at which 

 the velocity of reaction is great. Tiiese parts will be acted upon 

 amidst the formation of WCI,. In a lamp filled with chlorine the 

 wire, when glowing only weakly, was really found to be soon 

 corroded, while W^6V« was condensed on the wall opposite to the 

 wires. When on the contrary the wire glowed at higher temperature 

 it remained nearly intact, while WCI, was precipitated on the wall 

 opposite to the supports. The diameter of the wire increased even, 

 after 5 minutes the electric resistance was decreased by 12'/,. This 

 can easily be explained. By the high conduction of heat of the gas 

 the wall of the bulb is heated. Tiie ITCV^ is vaporized and dissociated 

 at the glowing wire, leaving behind on it a small quantity of tungsten, 

 while the chlorine is again liberated. The wire thickens at the 

 expense of those parts of the lamp that have a temperature at which 

 the velocity of reaction is great and the lamp is soon destroyed. 



Thus the chlorine describes a kind of cycle and we may ask 

 whether it would not be possible to introduce much less chlorine 

 into the lamp, so little, that those parts that have the temperature 

 of the maximum velocity of reaction are not destroyed after a few 

 hours, but only after nearly 1000 hours by the action of the chlorine. 



When however the pressure of the gas is lowered, the conduction 

 of heat decreases, the glass wall becomes less waim, the Tr6V„ that 

 has been formed is no longer vaporized and no chlorine is formed 

 back. In fact it was proved that in this way nothing could be 

 reached. After a short time all the chlorine in the bulb was fixed 

 and lamps with small chlorine fillings behaved after a short time 

 as if not any chlorine was present. 



The other halogens and oxygen behave in a similar way. By 

 the low vapour pressure of WOj there is here still less question of 

 a cycle. 



Better results were reached by the introduction of substances into 

 the lamp, which during the bujning develop continually small quanti- 

 ties of chlorine or oxygen. F. Skaupy was the first who suggested 

 this method. A small quantity of potassium-thallium-chloride is 

 introduced in a tube in such a way into the lamp, that during the 

 burning its temperature becomes high enough to develop chlorine. 



