570 



§ 9. 4. Anodic light emission has often been observed <iii ring electro- 

 lyses, when an insoluble or sparingly soluble reaction product is 

 formed at the anode. This product can then form either a solid 

 layer firmly attached to the anode, or a layer that gets more or less 

 easily detached. 



The former can often be observed in valve cells. Already below 

 the limiting tension a dullish white light may be seen at the valve 

 anode (10), which becomes |)retty intense under definite circumstan- 

 ces, (e. g. with Al-anodes in borax solution, Ta in diluted alkali or 

 carbonate solution). With this emission of light should also be 

 classed the emission of light of magnesium anodes in diluted 

 alkali (15). 



In all these cases the potential rises to the maximum value avail- 

 able, the passage of the current is almost entirely prevented. The 

 luminescence begins very soon after the closure of the current, often 

 with periodic oscillations of the intensity during the fiist minutes, 

 and then continues to persist till llie current is broken. The light 

 enussion is, however, generally soon jirevented, when electrode or 

 electrolyte are heated by the weak current that continues to pass. 

 In prolonged experiments it is, therefore, necessary to ensure good 

 cooling. 



The light, which is almost alwa_ys a dullish white, sometimes more 

 greenish or bluish, appeared to be continuous on spectroscopic in- 

 vestigation. 



It is also noteworthy that with magnesium anodes the maximun* 

 of light intensity is reached in potassium hydroxide J :100; a very 

 strong luminescence is also obtained by using an ammoniac solution 

 of di-sodium phosphate instead of the hydroxide. In this medium also 

 zinc anodes produce an exceedingly beautiful light emission, a borax 

 solution being the most suitable electrolyte with alumininm. But 

 also with aluminium and with tantalum diluted alkali hydroxide 

 solution etc. can be used. 



In these processes the electrolyte is covered by an adhering layer 

 of the oxide or of another insoluble anode product, as this was 

 already shown by other experimenters. The generality of such 

 phenomena is brought out by the fact that always new observations 

 of the kind described are being communicated (cf. e.g. la). 



But also when no direct valve actions are to be observed, such 

 phenomena of light are nearly always found when at the anode a 

 sparingly soluble product is found. To these belong, among others, the 

 following phenomena of luminescence which have partly already 

 been known for some time: 



I 



