348 Professor Liveing, On Differences 



luminescence of vapours of potassium, zinc, and magnesium, are 

 reversible. But perhaps the most striking case is that of the 

 cyanogen bands. The two sets of these bands which are con- 

 spicuous in the kathode glow in cyanogen are invariably associated 

 with two other sets, ultra-violet, near the solar lines L and N 

 respectively. Of these that near L is reversed when the flash of 

 an arc is taken in an atmosphere of cyanogen, as Professor Dewar 

 and I found 1 . 



Wood and Moore 8 have recently obtained, as a fluorescence of 

 sodium vapour in sun-light, a bright banded spectrum correspond- 

 ing to the absorption spectrum of the same vapour described by 

 Roscoe and Schuster 3 . I have little doubt that the kathode raj^s 

 would produce a similar spectrum in sodium vapour of sufficient 

 density, but could not try it because my glass tubes would give 

 way at the temperature required to produce such dense vapour. 



It is true that the oxygen lines and bands of the kathode 

 glow have not been observed to be reversed, whereas some other 

 absorption bands of oxygen are well known. I am not inclined to 

 attach much weight to the former fact, and with regard to the 

 latter it must be borne in mind that it requires a great thickness 

 of compressed oxygen to make those absorptions visible, so that 

 we could not expect to see the corresponding emissions from a 

 small mass of gas at low pressure. 



It is noteworthy that the persistent glow of radium salts in 

 atmospheric air is that of the kathode in nitrogen, as Sir W. and 

 Lady Huggins have shewn 4 . 



The behaviour of the metals suggests that the positive ions 

 consist of molecules of the vapour, which when deionized vibrate 

 to the same tune as do the like molecules under the influence of 

 kathode rays. In this connexion the monatomic character of 

 metallic vapours may have some significance. But be that as it 

 may, the positive ions may very well be molecules of the vapour, 

 and no reason occurs to me why chemically compound molecules, 

 such as those of carbonic oxide, should not become positive ions. 

 In fact these molecules often play the part of elementary atoms 

 in chemical combination. 



Of hydrochloric acid no emission spectrum has, so far as 

 I am aware, been observed, and the only absorption spectrum it 

 gives is a continuous one in the upper region of the ultra-violet, 

 so that if it fluoresces at all the light it emits may be in that 

 region. It is equally probable that the ions it furnishes may be 

 derived from it by chemical dissociation. The absorption by water 

 vapour is chiefly in the low red, and is only visible when the light 

 has traversed a great thickness of vapour, while its emission 



1 Proc. B. S. xtav. 247, note. 2 Astro-Phys. J. xvni. 94. 



3 Proc. II. S, xxn, 362 4 Proc, B. S. lxxji. 409, 



