571 



Electrolyte 



Phenomenon 



KJ aq, saturated 

 H2SO4 cone. 



KOH aq, strong 



NajHPOj NHjaq 



a bright luminescence at anodes of Cd, Hg, Pb. 

 « „ " . » .... Pb, Al, Ta; Mg gives a 



short flash; at Cd anodes there is seen a ring of light, which 

 moves up and down. 



Fe (a bright luminescence, but which cannot very easily be 

 examined on account of strong foaming), Ni (very slight 

 intensity of the current). 



Cu gives a circle of sparks. 



Exceedingly intense is the Inminescence at an Hg-anode in saturated 

 Kl-solution at snflicient density of tlie current. Tiie bright anode- 

 surface is covered with a thin layer of mercury iodide iinraediately 

 after tlie closure of the current, and then begins to emit a golden 

 light. After a short lime Ihe intensity of this light reaches a maximum, 

 and then diminishes again. By renewal of the mercury surface, 

 either by stirring or by allowing the mercury to overflow from a 

 funnel-shaped anode vessel, etc. the luminescence can be restored 

 with full intensit}'. 



In agreement with former experimenters (2) the spectrum of the 

 emitted light was found to be continuous, with a maximum of the 

 intensity in a definite spectral region. Wilkin.son (2) has pointed 

 out that the colour of this light also agrees with that of the light 

 emitted by the anode product in question, wlien it is bombarded 

 by cathode rays. 



^ 10. It is exactly these kinds of luminescence that are very often 

 considered as reaction luminescence (chemi-luminescence). Formation 

 or decomposition of the anode products were thought to be accompanied 

 by a luminescence which could reach a considerable intensity with 

 sufficient reaction velocity '). Bancroft (1) and his pupils, also 

 Wilkinson (2) have endeavoured to give support to this view. In 

 the course of our own observations on comparisjon with those of 

 other investigators it appears, however, that this conception is untenable. 



In the first place it can be established that all the phenomena 

 described in this chapter, are related. And this not only because 

 they appear to be of the same nature spectroscopically, but also 

 because their occurrence always appears to be bound to the formation 

 of sparingly soluble or unsoluble anode products. 



^) On this conception compare (5). 



