102 Prof. H. E. Armstrong. On Chemical Change, [Mar. 13, 



No attempt has been made as yet to deal practically with the 

 problem of electrical discharge in gases in the way in which Dixon, 

 Brereton Baker and Shenstone have dealt with that of chemical 

 change. Vacuum tubes are invariably made of soft glass — which is 

 altogether unfitted for accurate experiments, as it is always more or 

 less readily attacked by moisture, yielding up traces of alkali; no 

 special care is taken in cleansing them; no special precautions are 

 used in filling them ; and both oil of vitriol and commercial phosphoric 

 anhydride are used as drying agents, although it is well known that 

 these may be fruitful sources of contamination. And the electrodes 

 offer special difficulties, which it may be impossible to overcome, owing 

 to the readiness with which metals occlude gases. 



If there be evidence to show that the discharge in vacuum tubes 

 is conditioned by the presence of impurities, as I contend there is, it 

 is open to question whether most, if not all, of the effects attributed 

 to the so-called radiant matter — or as Sir Wm. Crookes now expresses 

 it, to electrons — are not in reality due to ordinary gross matter. To 

 take an example from this author's recent communication to the 

 Society on " Eadio-activity and the Electron Theory," at the close of 

 the paper reference is made to the discharge from a silver pole of 

 electrons which cause the glass against which they strike to glow ; at 

 the same time, the silver volatilises and is deposited near the pole. 

 " While the volatilisation of the pole is rapidly proceeding, the metal 

 glows as if red hot. This '.red heat ' is superficial only." It appears 

 to me possible to give a more ordinary explanation of the remarkable 

 phenomenon thus described by Sir Wm. Crookes. A silver electrode 

 would be more or less "polarised" with oxygen — for we know that 

 silver has the power of absorbing oxygen when fused, and that the 

 gas is only partially extruded as the metal cools ; in any case a 

 minute amount of oxygen would be present in the tube. When heated 

 by the discharge in vacuo the silver-oxygen "compound" would 



experiment exhibited by Dewar at the Royal Institution, showing that the 

 discharge at once ceases in one of Crookes' phosphorescent tubes on cooling with 

 liquid oxygen. Such treatment cannot be supposed to cause the condensation of 

 the residual air, but may well condition the deposition of the traces of vapour 

 (? of conducting water) present in the gas, thereby destroying the systems within 

 -which conduction can alone occur. 



It may be here pointed out that this argument is perhaps also at least partially 

 applicable in explanation of the erratic behaviour of the discharge in vacuum 

 tubes. If the discharge occur within a complex system, it may well be that 

 certain conditions will favour the formation of a conducting system including one, 

 and others of a system including another, of several substances present in a tube 

 — such an explanation, for example, would account for the fact that the mercury 

 spectrum is only sometimes seen in tubes connected with a mercury pump. And 

 having regard to the fact that a gas may exhibit several spectra, it may be that 

 the spectrum of a given substance varies more or less according as it is included 

 in one or other of several conducting s\ stems. — (' Chem. Soc. Trans.,' 1895, 1141.) 



