December 31, 1914] 



NATURE 



495 



is no longer coloured by kathode rays (or only in a 

 ven,- slight degree, indicating minimal traces of 

 sodium chloride). But there are other preparations 

 which, so far as I know, cannot be acquired in pure 

 condition by any means, not even by fractional crystal- 

 lisation. I never came across a pure sodium sulphate 

 • — the purity exists only on the manufacturers' labels. 

 Even the best preparations of this salt contain an 

 amount of sodium carbonate which up to the present 

 cannot be separated from it, not even by frequent 

 fractional crystallisation. The colour produced by the 

 small admixture, which always remains, is a very- 

 marked ash-grey. By an intentional further addition 

 of sodium carbonate the colour becomes nearly black. 



The question arises : What may be the cause of 

 these colorations in pure salts and also in solid solu- 

 tions of them? Shortly after the colours of the alkali 

 salts had been discovered, an explanation was given, ^ 

 according to which the phenomenon mainly consists 

 in a chemical reduction. For instance, in the case of 

 potassium chloride the chlorine wou'.d be set free, 

 while the remaining potassium is dissolved in the un- 

 altered main quantity of the salt, colouring it at the 

 same time. And it- seemed a convincing proof for this 

 theory when Giesel * and also Kreutz, simplv bv heat- 

 ing rock salt in the vapours of sodium or of potassium, 

 produced colours in this rock salt quite similar to those 

 produced by kathode rays. It seemed that the problem 

 was settled finally. However, it was soon discovered 

 that the coloured Giesel salts, although they look to 

 the e)'e quite like the kathode-ray salts, in all other 

 respects behave quite differently. For instance : — 



(i) The kathode-ray salts, as I mentioned before, 

 are very sensitive to daylight : after an exposure to 

 diffuse daylight of a few minutes — or in some salts 

 even of several seconds only — the coloration 

 diminishes, whilst the Giesel salts remain unaltered 

 even when they are kept in full sunshine for days or 

 even weeks. 



(2) The kathode-raj- salts, if dissolved in distilled 

 water, show absolute neutral reaction ; the Giesel salts 

 are strongly alkaline. 



(3) The kathode-ray salts give very marked photo- 

 electric effects (as Elster and Geitel ^ observed) ; the 

 Giesel salts are quite ineffective. 



(4) In certain circumstances, which will be men- 

 tioned further on, the kathode-ray salts may emit a 

 phosphorescent light, the Giesel salts none at all. 

 Therefore the question arose again. Whether there is 

 not a marked internal difference between the kathode- 

 ray salts and the Giesel salts, and what is the nature 

 of the latter? 



I have succeeded in settling this question, having 

 produced salts hy kathode rays, the behaviour of 

 which is in every respect absolutely identical with that 

 of the Giesel salts. You may produce such substances 

 if you allow- the kathode rays to fall on the original 

 salts not for a short moment only, but for a somew hat 

 prolonged time, until the salts are strongly heated. 

 Produced in this way the salts will keep colours ; but 

 the substances coloured in this way are not sensitive 

 to light ; they show no photo-electric effect ; they give 

 strong alkaline reaction, and they are not suited for 

 phosphorescence — all like the Giesel salts. It is quite 

 sure, and you may test it also directly by spectroscopic 

 proof, that in this case, if for instance you have 

 worked on sodium chloride, the chlorine is set free. 

 Then, of course, an amount of free sodium is left, 

 which dissolves itself in a deeper layer of unaltered 

 sodium chloride, to which the kathode rays could not 

 penetrate. I call these non-sensitive colours the after- 



3 E. Wiedemann and O. C. Schmidt, »7«-<i Ahh., liv., 6tS. 



■» F. Gifsel, Ber: D. Ckem. Ces., xxx., 156. 



S J. Elster anH H. Geit-1, ll'ieif. Ann., lix., 487. 



NO. 2357, VOL. 94] 



colours of the second class, while the ordinary sensitive 

 after-colours, produced in a short time on cool salts, 

 are called after-colours of the first class. 



Now-, if the after-colours of the second class are 

 identical w-ith those of the Giesel salts, then, of course, 

 the very different substances of the first class cannot 

 be also identical with the Giesel salts. Therefore thi 

 question arises anew what is the nature of the first- 

 class after-colours? 



One observes with regard to solid solutions that the 

 first-class colours depend not only upon the metal 

 contained in the small admixture, but thev var\ 

 greatly, for instance, in the case of the admixtun 

 consisting of potassium chloride or bromide or iodidt 

 This indicates that the metals alone do not cause th 

 after-colours. It becomes much more clear when we 

 expose some ammonium salts to the kathode rays. 

 (The ammonium salts are cooled by liquid air in the 

 discharge-tube to prevent their evaporation.) Then 

 you get strongly marked after-colours likewise ; for 

 instance, ammonium chloride becomes yellow-greenish, 

 the bromide becom.es yellow-brown, the iodide become- 

 brown, and the fluoride a deep blue. In the daylight 

 these colours are gradually destroyed, quite like other 

 after-colours of the first class. The colours them- 

 selves — yellow-greenish for the chloride, yellow-brown 

 for the bromide, and so on — induce us to presume that 

 the after-colours in this case are produced by the 

 haloids, and not by the hypothetical ammonium 

 radicle. This presumption becomes a strong convic- 

 tion when we observe that also a great number of 

 organic preparations which contain no metal at all 

 (and not any metal-like radicle) acquire marked after- 

 colours of the first class in the kathode rays also. 

 (The part of the discharge-tube which contains the 

 organic substances is cooled by liquid air.) 



Then you may observe that solid acetic acid 

 (C,H402) remains quite colourless in the kathode rays ; 

 but if you substitute a hydrogen atom by chlorine, 

 the substance thus produced (the monochloro-acetic 

 acid) acquires a marked yellow-green after-coiour. If 

 you introduce an atom of bromine instead of chlorine, 

 you get CjHjBrOj, and the after-colour is of a marked 

 yellow. Bromoform (CHBr,) turns into the colour of 

 loam, and chloral (CjHCljO) becomes a deep yellow. 

 In this way we see that not only salts, but likewise 

 substituted acids, substituted hydrocarbons, and sub- 

 stituted aldehydes acquire after-colours if they contain 

 any haloid. 



Now, it seems highly improbable that in the case of 

 alkali salts the electro-positive component is absorbed 

 only (producing the after-colour), and that, on the 

 other hand, in the ammonium salts and in the organic 

 substances the electro-negative component is efficient 

 only. The most probable inference is that in each 

 case both components remain and that both are 

 efficient, but that under the same conditions the 

 haloids produce a slighter colour than the metals, so 

 that in the case of the salts the haloid colour is over- 

 whelmed by the metal colour. 



Therefore we are compelled to suppose that we have 

 not to deal with a decomposition in the ordinarj- form, 

 by which the different components are finally separated 

 from each other and at least one of them is set entirelv 

 free, but that the components detained by absorption 

 remain at a quite short distance from each other, so 

 that they may easily meet again. I realise that — for 

 instance, in the case of sodium chloride — at everv 

 point of the coloured layer there is an atom (or per- 

 haps a molecule) of chlorine and an atom (or a mole- 

 cule) of sodium ; but they cannot combine, because they 

 are fixed by absorption and distended from each other by 

 the absorptive power, which in this case surpasses the 

 chemical affinity. But the absorptive pow-er may be 



