1075 



metal will be formed, which will not be less active than the alcaline- 

 metal. 



Finally we may ask, whether not many more substances possess 

 the "phosphorns-replacing" action. To try this lamps were constructed 

 in which the incandescent body had been sprayed with difïei'ent 

 oxides. We tried ZrO„ ThO,, TiO,, SiO„ CaO, Na,0. 



With all a strong glow-discharge took place at the tii'st burning. 

 The SiO^ only made an exception. Though the activity was less 

 than in the case of the glass-powder, the silicium-dioxide gave satis- 

 factory results. 



This suggests, that also the SiO^ is decomposed a little tirst, so 

 that silicium is liberated, which can replace the piiosphorus. Later 

 on the Si0.j on the wall can of course prevent again the blackening 

 (comp. § 3). 



The percentage of Si O.^ however that is decomposed in the 

 "sprayed lamp" is very small compared with the principal mass of 

 this substance, which is vaporized without ha\'ing been decomposed, 

 then condensed on the wall, where it prevents the blackening. 



In order to learn more about ttiis we have drawn out quartz-tubes to capillaries 

 and slipped these over the tungsten wire. When now the tungsten wire is glowed 

 in vacuo at a high temperature, a relatively slow vaporization of the Si02 takes 

 place. Finally however a considerable precipitate is obtained, which strongly shows 

 the interference colours. When the bulb is opened, gradual changes are observed 

 of the quartz which forms such a thin layer. Everywhere needles are formed 

 stretched in one direction, which also show interference-colours, but in which already 

 with the naked eye irregularities are seen; with 50-fold magnification already a 

 shell-shaped conglomeration-structure of those needles is observed. Sometimes such 

 an alteration remains very local; doubtlessly it is due to the difference in dilata- 

 tion-coefficient of the glass (under layer) and the quartz (upper layer) ')• 



In a following series of experiments the temperature coefficient of the resistance 

 of the wire was investigated to find out whether it had taken up some Si at the 

 supposed decomposition of the Si02. 



It will be known, that the temperature-coefficient of the resistance offers an 

 extremely sensible method to see whether an element contains any impurity. 

 Tungsten and silicium allie very easily, so that the formation of Si must become 

 evident from the change of the temperature-coefficient of the resistance. 



Only very little silicium however is needed to remove rest-gases (with the aid 

 of glow-discharges). Let us suppose for the sake of simplicity, that the only rest- 

 gas that occurs is oxygen with a total pressure of 0,006 mm ; in a bulb with a 

 volume of 150 cm^ this would correspond with 0,0017 mg 0., which could only 

 be bound by 0,0015 mg Si. 



Little silicium is needed not only, but little silicium is formed also in the lamp. 

 That is why the method of the temperature-coefficient measurement does not give 

 a positive indication on the formation of Si; this is also due to the fact that the 



M Also R. B. SosMAN (Journ. Ind. and Engin, Ghem. 8, 985 (1916) stales at 

 condensation of quartz-vapour a pronounced inclhiation to form needles of amorphous 

 quartz. 



