( 347) 



Wlien heating the sodiuiii for the tirst time a pretty large quan- 

 tity of gas escaped from it (according to Wood hydrogen), whicii of 

 course was pumped off. After the apparatus had functionated a 

 couple of times, the tension within the tube remained for weeks 

 less than 1 mm. of mercury, also during the heating, which, in the 

 experiments described in this paper, nexev went beyond 450°. The 

 inner wall of A^X' and also the small tubes A and B are after a 

 short time covered with a layer of condensed sodium, which favours 

 the homogeneous development of the vapour in subsequent heatings. 

 It is remarkable that scarcely any sodium condenses ou the parts of 

 the tube that stick out of the furnace, so that also the windows 

 remain perfectly clear. The density of saturated sodium vapour at 

 temperatures between 368° and 420° has been experimentally deter- 

 mined by Jewett ^). He gives the following table. 



These densities are of the same order of magnitude as those of 

 mercury vapour between 70^ and 120°. At 387° the density of 



^) F. B. Jewett, A new Method of determining the Vapour-Density of Metallic 

 Vapours, and an Experimental Application to the Cases of Sodium and Mercury. 

 Phil. Mag. [6J, 4, p. 546. (1902). 



23 



Proceedings Royal Acad. Amsterdam. Vol. IX. 



