[ 331 ] 



XL. Intelligence and Miscellaneous Articles. 



ON THE SPECTRA OF NITROGEN AND THE ALKALI-METALS IN 

 GEISSLER TUBES. BY G. SALET. 



lyr SCHUSTER made public, in 1872, the important fact that ni- 

 -'-'-** • trogen, heated in a Greissler tube with metallic sodium, no 

 longer gives its characteristic grooved spectrum. He described the 

 bright lines obtained in this case, and attributed them to pure ni- 

 trogen — the band spectrum being, in his opinion, that of an oxide of 

 nitrogen, a compound which is destroyed by the alkali-metal. 

 Later, doubts arose concerning the validity of these conclusions ; 

 for, on the experiments being repeated, the grooves were indeed 

 seen to disappear after the action of the sodium, but they were 

 replaced by various spectra, of which not one really belongs to 

 nitrogen, so that after purification this gas could no longer be 

 detected by the prismatic analysis. I have also remarked that the 

 chemical compound which is in reality formed by the action of 

 oxygen upon nitrogen is the peroxide, a very stable body and one 

 the spectrum of which does not at all coincide with that of which 

 we have to account for the appearance. 



My present purpose is, to demonstrate (1) that the grooved 

 spectrum can be produced with nitrogen heated in contact with 

 sodium, (2) that the disappearance of the spectrum of nitrogen is 

 due to the disappearance of the nitrogen itself, it being entirely 

 absorbed by the sodium under the influence of the electric effluvium, 

 and (3) that the spectrum described by M. Schuster is very pro- 

 bably to be attributed to the vapour of the alkali-metal. 



1. An account of the series of experiments which have led me 

 to these conclusions would be too long to give here ; I will only 

 cite the most decisive. I had a tube of hard glass blown by M. 

 Alvergniat, stopped at both ends, 12 centims. in length and 2 

 centims. in diameter. Two aluminium electrodes were arranged 

 at one end of the tube, separated by a distance of about a centi- 

 metre ; to the other extremity a tubule with a bulb was soldered. 

 Into the bulb a little piece of sodium was introduced; and the 

 tubule was then cemented to the mercury air-pump. A vacuum 

 having been produced, the sodium was heated ; it swelled and 

 boiled for a long time, losing hydrogen ; it then ceased to boil, and 

 at a higher temperature slowly volatilized. The apparatus was 

 now severed from the air-pump by a stroke with a pipe, and the 

 shining globule of liquid sodium was conveyed into the experiment- 

 tube. After cooling, the bulk was separated, and the tube was 

 cemented directly to the pump. Exhaustion was then recommenced, 

 the sodium volatilized, care being taken that the condensation of 

 the metallic vapour should only be produced in that half of the 

 tube which carried no electrodes ; and nitrogen was admitted pure 

 and dry. I again exhausted, to the amount of three fourths of 

 the nitrogen, taking care each time to volatilize the alkali-metal, 

 and finally closed the apparatus, leaving in it a pressure of about 

 5 millims. I could then melt the globules, cause them to unite, 



