48C> 



Profs. Liveing and Dewar. 



[Dec. 11, 



For convenience in introducing the sodium, a fine platinum tube, 

 about 1 millim. in diameter, was filled with sodium by sucking up the 

 melted metal, and different lengths of this tube were cut off as required, 

 wrapped at once in thin platinum foil, and dropped into the heated 

 vessel. Amalgams containing various percentages of sodium were 

 also employed. In the case of amalgams containing 1 per cent, and 

 upwards the pieces were weighed ; those containing less than 1 per 

 cent, were measured by a very fine pipetbe. In all cases the amalgam 

 was introduced into the hot vessel wrapped in platinum foil. 



For varying the pressure within the vessels, and so by compression 

 or expansion varying the density and thickness of the stratum of 

 metallic vapour under observation, an ordinary compressing and 

 exhausting syringe, connected with a gauge, was used. With the 

 bottle we could not venture to reduce the pressure below that of the 

 atmosphere, lest the softened platinum should collapse ; but in the 

 case of the tube the pressure could be reduced to one-fourth of an 

 atmosphere. Into both vessels we have pumped nitrogen until the 

 pressure reached three atmospheres. 



Whenever it was desired to empty the vessel of sodium, in order to 

 begin a fresh series of experiments without the necessity of cooling the 

 vessel, the sodium (or sodium and mercury) vapours were blown out 

 by a current of hydrogen passed down to the bottom of the vessel in 

 the mode already described, and when the metallic vapours were suffi- 

 ciently expelled, the hydrogen was replaced by nitrogen. In this 

 way, though the last traces of sodium were never expelled, the D 

 absorption could be reduced to two narrow lines. 



On introducing a fragment of sodium into the hot vessel, when 

 free, or nearly so, from sodium to begin with, the bright yellow band 

 rapidly expanding and becoming fainter and more diffuse, and finally 

 dying away, was seen as in the iron bottles ; but instead of the D 

 absorption remaining narrow as it had been before, it began to in- 

 crease in width as the bright band died away, and soon attained 

 a maximum width, and remained steadily at that width — a width 

 several times that oi the distance between the D lines, but much less 

 than that of the bright band at its widest. A second piece of sodium 

 thrown in produced a repetition of the same phenomena, except that 

 the bright band was obscured by the absorption of the layer of 

 sodium vapour first introduced, and only seen when it expanded beyond 

 the absorption band. More pieces of sodium produced like effects ; the 

 absorption band gradually widening as more sodium was put in, until 

 the channelled spectrum began to appear. 



The effects of compressing the vapour were very remarkable. As 

 the pressure increased the channelled spectrum speedily disappeared, 

 then the diffuse edges of the D band contracted, the band itself 

 likewise contracting until it became a very fine pair of lines, 



