102 Lord Kayleigh — Spectroscopic Notes 



three-way tap. It may be argued that cnpric oxide is not 

 competent in moderate length to remove the last traces of 

 hydrogen from air, even though the air be passed over it in 

 a slow stream. 1 found, however, on a former occasion * 

 that hydrogen purposely introduced into nitrogen could be so- 

 far removed in this way that the weight remained sensibly 

 unaffected, although ^^ of residual hydrogen might be 

 expected to manifest itself. 



Moreover, when air purposely contaminated as above with 

 ^ of hydrogen was passed over the copper oxide, the 

 additional hydrogen appeared to be removed, the visibility of 

 C reducing itself to that corresponding to untreated air. 



Being desirous of testing the matter as far as possible, I 

 have experimented also with nitrous oxide and with oxygen. 

 In the former case the general appearance of the spectrum is 

 much the same as with air. The C-line was thought to be, 

 if anything, more visible than in the case of air, but the 

 difference could not be depended upon. Two samples of gas 

 were tried, one from an iron bottle as supplied commercially,, 

 the other prepared in the laboratory from ammonium nitrate- 

 Oxygen from permanganate of potash also showed the C-line 

 more distinctly than air, but this may probably be attributed 

 to the elimination of a neighbouring nitrogen line. It is 

 possible, of course, that these gases may have contained traces 

 of hydrogen, but in that case it is strange that the proportion 

 should be so nearly the same as in air. 



These observations certainly seem to leave a minimum of 

 room for the hydrogen found by M. Gautier, but I should be 

 unwilling to call his conclusion in question on the strength 

 of what are after all but eye estimates. I have not been able 

 to find a detailed account of M. Gautier's experiments or of 

 what precautions he took to assure himself that the water 

 collected could not have had its origin in the glass or copper 

 oxide of his hot tubes. The most satisfactory test would be 

 comparison experiments in which oxygen or nitrous oxide is 

 substituted for air, or, perhaps better still, in which air is 

 used over and over again. 



If, as I should suppose were I to judge from my own 

 experiments alone, the residual C-line was not wholly or even 

 principally due to hydrogen in the air, it would have to be 

 explained by hydrogen evolved from the glass of the sparking- 

 chamber or from the platinum electrodes. In view of what 

 is known respecting the behaviour of vacuum-tubes, such an 

 explanation does not appear improbable. 



* On an Anomaly encountered in Determinations of the Density of 

 Nitrogen Gas, Proc. Eoy. Soc. vol. 55, p. 343 (1894). 



