248 



Profs. G. D. Liveing and J. Dewar. [May 31, 



arc from a De Meritens' machine in hydrogen or nitrogen. It does 

 not show in the spark between magnesium electrodes of an induction 

 coil used in the ordinary way, either with or without a Leyden jar, in 

 hydrogen or in air at atmospheric pressure ; nor in the glow discharge 

 in vacuous tubes with magnesium electrodes when the residual gas is 

 either air, oxygen, hydrogen, carbonic acid gas, or cyanogen. Nor 

 does it appear, except in the one instance above mentioned, in the glow 

 discharge in highly rarefied air in a tube containing either magnesia 

 or a strip of metallic magnesium. 



A review of all the circumstances under which the triplet near 

 M and its associated bands appear, and of those under which they 

 fail to appear, leads pretty conclusively to the inference that they are 

 due not to merely heated magnesium but to the oxide, or to vibra- 

 tions set up by the process of oxidation. 



With reference to this triplet, Mr. Lockyer (he. cit., p. 122) has 

 referred to us as his authority for the statement that at the tempera- 

 ture of a Bunsen burner as ordinarily employed the ultra-violet line 

 visible is that at 373. We do not agree to this as a statement of 

 observed fact, and we cannot imagine how the passage to which 

 Mr. Lockyer refers (' Roy. Soc. Proc.,' vol. 32, p. 202) can be sup- 

 posed to warrant it. The flame we mention in that passage is not 

 that of a Bunsen burner but that of burning magnesium, which may- 

 be very different from the former even when the magnesium is 

 burning in the air which is mixed with coal gas in the Bunsen burner. 

 Moreover, whatever the temperature of the flame may be, we have 

 never observed the triplet at \3730 unaccompanied by other ultra- 

 violet lines. In the flame of burning magnesium, as we state {loc. 

 cit, p. 189), " photographs show, besides, the well-known triplet in 

 the ultra-violet between the solar lines K and L sharply defined, and 

 the line for which Cornu has found the wave-length 2850 very much 

 expanded and strongly reversed." 



We have expended a vast amount of time and trouble over 

 vacuous tubes, and our later experiments do but confirm the opinion 

 which we had previously formed that there is an uncertainty about 

 them, their contents and condition, which makes us distrustful of 

 conclusions which depend on them. Photographs of the ultra-violet 

 spectra given by such tubes tell tales of impurities as unexpected as 

 they are difficult to avoid. Every tube of hydrogen which we have 

 examined exhibits the water spectrum more or less, even if metallic 

 sodium has been heated in the tube or the gas dried by prolonged 

 contact with phosphoric oxide. Indeed the only tubes which do not 

 show the water spectrum have been filled with gases from anhydrous 

 materials contained in a part of the tube itself ; and even when tubes 

 have been filled with carbonic acid gas from previously fused sodium 

 carbonate and boracic anhydride the water spectrum is hardly ever 



