On definite Proportions. 405 



stance ? Or, can so small a quantity of quicksilver be suf- 

 ficient to disguise so completely the properties of potas- 

 sium ? Or, are both substances in the form of protoxides } 

 When I had kept a portion of this substance about a 

 month in a small stopped bottle, I found it surrounded by 

 a gray brown cracked crust, in the middle of which a centre 

 of the amalgam of potassium was found, containing so 

 much quicksilver that it was completely fluid. I took off 

 the gray brown crust, and threw it into water, in which it 

 occasioned a very brisk extrication of gas. When moisten- 

 ed with a drop of water, it evolved hydrogen with the 

 greatest violence, with heat and smoking. The water con- 

 tained potass, and left behind yellow oxide of mercury. 

 The gray brown crust therefore contained again a combi- 

 nation of potassium with quicksilver, in a condition of 

 which I cannot form a distinct idea. During the most 

 violent extrication of hydrogen gas, the quicksilver exhi- 

 bited itself in the highest degree of oxidation. Can this 

 be explained by an electrochemical polarity within the 

 fluid ? I think scarcely ; for the effect is the same upon 

 glass, as upon platina or wood. 



Since, according to Davy's account, a greater heat, than 

 that in which I distilled the amalgam of potassium, destroys 

 and perforates the glass, I give up the hope of obtaining- 

 potassium pure by means of my electrical battery. 



An unsuccessful attempt to separate the base of ammonia 

 from a boiling solution of sal ammoniac by means of Rose's 

 or D'Arcet's fusible mixture of bismuth with zinc and tin, 

 induced jne to attempt to collect potassium by means of 

 the same compound. I had hoped, that if the former base 

 could be collected in it, it would be easier to separate it 

 from water when cold, and to distil it, than the amalgam 

 with quicksilver; but in this hope I was disappointed. I 

 employed in this experiment a glass lube, in the lower end 

 of which a platina wire was cemented. On this wire I put 

 the rnetallic compound, I poured on it the concentrated 

 alkaline solution, and melted the metal by means of a lamp, 

 which was kept burning throughout the experiment. The 

 battery acted powerfully, and a considerable extrication of 

 gas took place, both from the metal and from the positive 

 wire. The solution became more and more saturated, and 

 after two or three hours it began to dry up. I now hastily 

 poured the liquid metal on a dry and cold saucer ; and it 

 immediately hardened. I wiped from its surface the potass 

 which adhered to it, and scraped off some particles of the 

 metal. Wheti laid on a piece of litmus paper, reddened by 

 C c » th« 



