565 



^ 3. In contrast with wliat was found by Morse, H-lines (espe- 

 cially R^) were present in the emitted speclrnni; further Pt-lines at 

 platinum cathodes. Apart frovi this it nms stated that electrolyte and 

 metal chromoscope, give totally different spectra, — a fact which was 

 quite overlooked both by v. Bolton and by Morse. When the metal 

 that is to be detected, only occurs in the electrolyte, the spectrum 

 very closely resembles that of the spark-spectrnm of the metal. If 

 this same metal is, however, immersed as cathode in pure acid, a 

 spectrum is obtained which agrees closely with the aic-spectrnm. 

 As an illustration of these facts, which 1 could verify repeatedly, 

 some photographs have been reproduced here ffig. 3). 



That we are justified in speaking of a general behaviour here, 

 follows for the rest, besides from our own observations (with Mg, 

 Pb, Fe, Wo, Mo, Ta, Al, Cu etc.), also from the data of v. Holton 

 and Morse themselves. If the metal is at the same time in electro- 

 lyte and electrode, it is to he expected that a superposition of the 

 two spectra is observed. Since, however, the metal chronioscopes 

 produce more intense phenomena, it is easy to understand that 

 Morse observed a strong arc-spectitim that is generally superposed 

 by a weak spark-spectrum. 



If the chronioscopes are to funclion normally, a definite current 

 intensity is required in both cases, which though dependent on the 

 adjustments of the apparatus, always remained within the limits 

 indicated by v. Bolton. With Cu-salt in the electrolyte chromoscope 

 (fig. I) e. g. 0,4 -0,5 Amp. appeared to be required. A greater 

 intensity of the current caused incandescence of the wire, and the 

 disappearance of the luminescence, whilst a- weaker current 

 caused the total light intensity to become smaller. As appears from 

 the adjoined photographs, also a selective weakening takes place: 

 some lines losing- much more in intensity than the rest. The same 

 effect may also he reached by greatly diminishing the concentration 

 of the metal salt. 



In many cases, especially when earth-alkali salts are used, one 

 has the impression that the whole liquid at the cathode is lumi- 

 nescent. This etfect is, however, not always found; besides the 

 spectrum was not changed by this. The co-luminescence seems to 

 be caused by still unknown accessory circumstances. 



With regard to the mechanism of the emission process it may be 

 considered as an established fact that the cathode is surrounded by 

 a gas envelope. As already Voller and Walter observed, this may 

 be shown simply as follows: when a well-luminescent chromoscope is 



