which becomes a violet-colourecl Gas ly Heat. 13 



find, a double decomposition, the chlorine quits the hydrogen 

 to unite to the potassium, and the substance quits the potassium 

 to unite to the hydrogen. 



The new substance, I find, is not decomposed when Voltaic 

 sparks are taken in it in its gaseous state from ignited points 

 of charcoal : at first there are white fumes, probably fiom the 

 action of moisture or hydrogen in cliarcoal, on the su!)stance ; 

 but these fumes soon cease, and when the tube in which the 

 experiment is made is cooled, the substance appears unaltered. 



From all the facts that have been stated, there is every 

 reason to consider this new substance as tin undecompounded 

 lody. In its specific gravity, lustre, the high number in wbich 

 it enters into combination and colour, it resembles the metals ; 

 but in all its chemical agencies it is more analogous to oxygen 

 and chlorine ; it is a non-conductor of electricity, and possesses, 

 like these bodies, the negative electrical energv with respect to 

 metals, inflammable and alkaline substances ; and hence, when 

 combined with these substances in aqueous solution and elec- 

 trized in the Voltaic circuit, it separates at the positive surface : 

 but it has a positive energy with respect to chlorine ; for, when 

 united to chlorine in the compound acid I have described in 

 pages 5 and 6, it separates from the chlorine at the negative sur- 

 face. This likewise corresponds with their relative attractive 

 energ}'. Chlorine expels the new substance from all its com- 

 binations on which I have made any experiments. 



The new substance seems to possess a stronger attraction for 

 most of the metals than oxygen ; but it is expelled from phos- 

 phorus and sulphur bv oxygen : I found by passing oxvgen and 

 the compound of it with phosphorus through a glass tube heated 

 red, phosphorous acid v.as formed, and the violet gas appeared. 



That it produces so little heat and so seldom light in entering 

 into combination, may be accounted for from its solid form and 

 its great weight as an element. Potassium, however, as I have 

 mentioned, burns in tire violet-coloured gas, and when this gas 

 is thrown upon the flame of hydrogen, it seems to su[)port it* 

 combustion. 



The saturating or neutralising powers of the new substance 

 appear to be greater than those of oxygen, and less than those 

 of chlorine. 



It agrees with chlorine and fluorine in forming acids with hy- 

 drogen, and it agrees with oxygen in forming an acid with 

 chlorine. 



In my first experiments I conceived that it formed substances 

 analogous to alkalies in combining with the alkaline metals ; 

 for the compound produced by its action upon solution of po- 

 tassa, even when the substance was in great excess, reddened 



turmeric 



