278 / On definite Proportions. ' 



any sulphur passed over. The nsass, when taken out, was 

 still in the form of a powder, only that its colour was 

 somewhat brighter, and it was still partly attracted by the 

 magnet. It was not however soluble in muriatic acid. 

 Two grammes of it were burnt in an open platina crucible, 

 and left 1-4 gr. of red oxide of iron not at all magnetic, 

 answering to -97 of metallic iron. Consequently the re- 

 maining ro3 gr. was sulphur; a\\i\ IGO pans of iron had 

 taken up 106-3 of sulphur. Since however the pyriies, 

 which was formed, was still partly mngnetical, I conjec- 

 tured that it somewhat resembled the red oxide, which by 

 too strono ignition has been reduced in a slight degree to 

 the state of a protoxide, and for this reason is again sub- 

 jected to the influence of the magnet. 



I iherefore distilled 20 gr. of very pure native pvrites in 

 a small glass retort with a rcce.ver. At first a trace ot 

 moisture passed over, which, when the experiment was con- 

 cluded, had asiumtd the lorm of pilv drops adhering to the 

 receiver, and which I took for concentrated sulphuric acid ; 

 but hv dilution with water this fluid became milk white, and 

 not at all acid. It was thtrefoie neither water nor sul- 

 phuric acid. I could not afford this substance any further 

 attention ; perha])S it w as the alcohol of sulphur. The 

 mass Icfi in 'he retort was exposed for some time to igni- 

 tion : it had lost 4*4 gr. of sulphur, which was collected in 

 the neck of the retort and in the receiver. Of the remaining 

 15*6 jr., live were dissolved in nitric acid: these being 

 evstporatetl to dryness, a d i<rnited, in a platina crucible, 

 left 4-3 of red oxide, which was not at all magntlical. 

 I>issoivtd in murianc acid, it kft -02 ol silica. Hence vve 

 have 1 1-416 of red oxide for ihe whole mass, or deducting 

 'OQ25- for silica, 13-33, answering to c)-i?3S of nreiallic iron. 

 There remain therefore iO-7 gr. for the sulphur. Conse- 

 quentlv 100 pans of iron had been combined with 115-5 

 of sulphur; that is, with nearly \\\\cq tlie quantity which 

 bad been found in the sulphuret at a nnniniuni. 



I repealed the experiment with some seieci pieces of an- 

 other specimen of pvrites. It was verv (jiiely powdired', • 

 roasted in ihe muffle of an assayiitg furnace, in a di'^h of 

 platina, and in tie mean lime occasionally stirred with a 

 hook of the same metal. Ten graiiuiits ot pyrites afforded 

 me 6-67 of red oxide, not in the hast magiiciicai, ieaving 

 •07 of silica when- dissolved in the nuiriatic acid. 1 hese 

 6ft er. of red oxide indicate 4-5773 gr. or' pure iron, which, 

 addled to the quantitv of silica, and suhira-.ted from lh« 

 whote weight, leaves 5-33'-25 for the sulpluir. Consequently 



loq 



