12 On some new analytical jResearches 



very soluble in hydrogen ; but, under common circiim- 

 st.-.nces, hvdrogtn docb not seem to be absorbable by po- 

 tassium. 



III. Analytical Experiments on Sulphur. 

 I have referred, on a former occasion *, to the expcrimenl3 



tnctal, and in ;he process tl'.cre is an alosorption both of hyi'.rogen aiid oxy- 

 pen. It is likewise produced in experiment; on the generation of potassium 

 b5' exposing potash to ignited iron, at the time (1 believe) that common air is 

 admitted, dutinj; liie cooling of tlie tube. 



It is non-conductip.g;, Inflanies spontaneously in air, and produces potash 

 and aqueous vapour by its combustion. 



When potr.ssiu^T! is heated in hydrogen in a flint g^as- retort, or even for a 

 great lentrth of tini.c in a green glass r.etort, there is an absorption of th^ gus; 

 but this is independent of the presence of potassium, and is ow-ing to the ac- 

 tion of the metallic oxides in ihe glass upon the hydrogen. 



If a solid compound of hvdrogen and potas'slum could he formed, ".vc might 

 expect its existence in the experiment with the gun-b.-irrel, in which potas- 

 sium is exposed to hydrogen at almost every temperature ; but tht metal 

 formed in this process, when proper precautions are taken to exclude car- 

 bonaceous matters, is uniform in its properties, and ge.;erates for equal quan- 

 titie';, egua! proportions of hydrogen by the action of water. 



The general pha:namena of this operation, sho-.v indeed that the solution 

 of potassium in hydrogen is intimately connected with the general principle 

 of tlie decomposition, and confirm my first idea of theaction ui the two bodies. 



Hydrogen dissolves a large quantity of potassium by heat, but the greater 

 portion is precipitated on cooling. The attractions which dciern>ine the che- 

 mical change, seem to be that of iron for oxygen, of iron for potassium, and 

 of hydrogen for potr^ssium ; and in experiments, in which a very intense heat 

 is tised for the production of potassium by iron, Ihave often found, that the 

 gas -.vhlch comes over, though it has pas.sed through a tube cooad by ice, 

 inflames spontaneously in the atmosphere, and burns with a most brilliant 

 ligj>t which is purple at the edges, and throws ofTa dense vapour containing 

 potash. 



Sodium appears to be almost insoluble in hyd.-ogcn, and this seems to be 

 one reason wiiy it cannot be obtained, except in very minute quantities, in 

 the experiment with the gun-barrel. 



Sodium, though scarcely capable of being dissolved in hydrogen alone, 

 seems to be soluble in ths compound of hydrogen and po'ai'^Ium. By exposing 

 mixtures of potash and soda to igil^tcd iron, I hav'c obtained some very cu- 

 rious alloys ; which, whether the potassium or the sodium was in excess, were 

 fluid at com.mon temperatures. The compound containing an excess of po- 

 tassium was even lighter than potassium (probably from its flr.idlty). All 

 these alloys -.verc in the highest degree inflammable. When a globule of the 

 fljid^alloy was touched by a gicbtileof mercury, tltcy combined with a heat 

 that singed the piper upon which the experiment was made, and formed, 

 when cool, a solid so hard, as not to be cut by a knife, 



» B:ikeriin Lecture, 1808, p. 16, 



of 



