572 



Premising this, it may be inferred from a pretty great number of 

 reasons tiiat a conce|»tion of tiiese luminescence phenomena as reaction 

 luminescence phenomena must l)e considered as erroneous. 



In the first place with this view of the matter it cannot be ex- 

 plained why only formation of unsoluble products gives rise to 

 luminescence. It can, indeed, be predicted that the probability of 

 anodic luminescence and its intensity will be the greater as the 

 anode-product dissolves with the greater diflicully. For it appears 

 in particular that on formation of readily soluble anode-products, 

 luminescence never seems be observed. 



Nor can the considerable increase of intensity of the luminescence 

 at low temperature (hence at smaller reaction velocity) be accounted 

 for on the ground of the said conception. For with valve anodes 

 ihe luminescence is by no means most pronounced on particularly 

 strong anode-reaction, but only when the excluding layer is as stable 

 and homogeneous as possible, and is attacked as little as possible 

 by the electrolyte. Thus magnesium emits the brightest light in 

 (h'hitfd KOU, aluminium in borax solution, which would certainly be 

 unaccountable in the case of real "cherai-luminescence". A mag- 

 nesium anode is particularly strongly attacked by diluted sulphuric 

 acid, though all the same, there is no luminescence at all to be 

 observed. 



Moreover it lemains inexplicable how anodes which are rapidly 

 covered by an insoluble layer, yet continue to emit light. It might 

 much sooner be exjiected in this case that the light would cease 

 after the formation of a covering layer. But this is by no means 

 observed in the majority of the cases. 



Finally the inciease of light intensity after the closui'e of the 

 current, as is particularly clearly observed with mercury anodes in 

 Kl-solution, is unaccountable in a reaction luminescence. For, how 

 a certain quantity of reaction products would be able to increase a 

 direct chemi-biminescence, is not clear. Nor can periodic and rhythmic 

 light emissions (Cd in Kl-solution, Mg immediately after the closure 

 of the current) be accounted for in this way. 



^ 11. The only conception which can be brought to harmonize 

 with all the experimental facts, is in contrast with the conception 

 discussed just now, in my opinion the following: at once after the 

 closure of the current a layer of reaction products is formed at the 

 anode, which hampers the passage of the ions to the anode, or 

 renders it impossible. Then the electric discharge of these ions takes 

 place (at sufficiently high potential) under the influence of split off 



I 



