1873.] 



On the Vapour -density of Potassium. 



203 



I. <c On the Vapour- density of Potassium/-' — Preliminary Notice. 

 By James Dewar and William Dittmar. Communicated 

 by H. E. Roscoe, F.R.S., Professor of Chemistry in Owens 

 College, Manchester. Received January 27, 1873. 



Since the elaborate experiments of Deville and Troost on the vapour- 

 densities of substances at high temperatures, little has been added to 

 chemical science in this field of research. Doubtless this is in great part 

 owing to the difficulty of any one student manipulating the complex 

 apparatus necessary for the execution of the experiments. But the 

 operations are greatly increased in difficulty when we select bodies that 

 are readily inflammable in air and attack with facility glass and porcelain 

 at the high temperatures to which they are exposed. This is the reason 

 why the molecular weights of a most important class of elementary 

 bodies, viz. the alkali metals (although these are volatile at moderate 

 temperatures), have remained to the present time undetermined. It was 

 with the view of adding something to our knowledge in this department, 

 that we recently undertook some experiments with potassium, the results 

 of which we now beg leave to lay before the Society. The special diffi- 

 culties we had to overcome are involved in the endeavour to answer the 

 following questions : — 



1. Is it possible to convert potassium into a gas of one atmosphere's 

 pressure at any of the constant temperatures we can at present command ? 



2. Is it possible to generate pure potassium-vapour and to keep it 

 from getting oxidized ? 



3. Supposing a definite volume of such vapour to have been procured, 

 how can its weight be ascertained ? 



After a succession of failures, which we shall not detail, we at last 

 succeeded in devising a workable process, which may be briefly described 

 as follows : — 



A cylindrical iron bottle of at least 200 cub. centims. capacity, of a 

 thickness in the body ensuring sufficient rigidity at even a bright red 

 heat, and provided with a well-ground inbent neck, pierced with a canal 

 of about 2 millims. in diameter, is employed as a generator and recep- 

 tacle of the vapour. 



A mass of about 20 kilogrs. of zinc contained in a plumbago crucible, 

 which being placed in a forge-fire can be readily heated up to the 

 boiling-point, serves as a bath. 



The experiment begins by first deoxidizing the inside of the receptacle 

 at a red heat by means of a current of dry hydrogen, which is continuously 

 .maintained until the bottle has cooled down belov/ redness, At this 

 stage about 200 grms. of pure mercury are introduced into the bottle, 

 which is then inserted into the red-hot zinc, without, however, covering 

 the upper extremity of the bottle. After f of the mercury is distilled 

 off (which is accomplished in a very short time), the neck is withdrawn", 



s2 



