1814.] On the Cause of Chemical Proportions. 255 



lower layer. I threw it on a filter, and after having well washed 

 it, 1 allowed it to dry. It resembled exactly dry perhydrateof iron. 

 1 reduced it to powder, and exposed it for 24 hours to a heat of 

 100° Fahrenheit. I found it to be a hydrate of rhodium, contain- 

 ing neither potash nor muriatic acid. I introduced 100 parts of 

 this hydrate into a small retort exactly weighed, and heated it over 

 a spirit lamp. INIy object was to drive oft" the water by a moderate 

 heat, and after having ascertained the diminution of weight occa- 

 sioned by its separation, to expel likewise t!ic excess of oxygen ia 

 the oxide. 1 obtained at first pure water; but the oxide having 

 been too much heated at the bottom of the crucible, appeared to 

 catch fire, and disengaged in a moment its excess of oxygen, and 

 left for residue a brittle greyish jnass with the metallic lustre, 

 whicii weighed "Ji parts. Thus the oxygen and water together 

 weighed 26 part''. The grey mass appeared at first to he metallic 

 rhodium, but wlien I mixed it with some drops of fat oil, and 

 heated it a little, a violent detonation tock place, and the rhodiimi 

 was reduced to tlie metallic state. 



It is evident from the properties of the oxide examined, (as for 

 example the colour of its salts, and its solubihty in muriatic acid 

 witliout disengaging oxymuriatic gas,) that it differs in its state of 

 oxydation from the two oxides above examined; and it appears 

 (juite clear, that it must contain more oxygen than they. But if 

 the two oxides above described be R + C), R 4- 2 O, this oxide 

 must be at least II + 3 O. 1 have proved that gold forms two sali- 

 fiable oxides, of which tlie one contains 3 times as much oxygen as 

 the otlier. 1 have likewise made it probable, that the purple of 

 Cassius contains an oxide not taliliable, intermediate between tlie 

 two others ; thut is to say, that the oxides Au 4- O, Au -f 3 O, 

 combine with acids, while the oxide Au + 20 does not combine 

 with these bodies, thougii it lias au affinity willi other oxides. Now 

 the same thing appears to hold with rhodium. The oxides Rh + 

 O, Rh + 3 O, form salts witli acids, wliile the oxide Rh + 20 

 combines only wlih alkaline bodies. If we calculate, according to 

 this supposition, the result of our analysis of the hydrate of rho- 

 dium, we find that /^ of l)rotoxide of rhodium ought to he equi- 

 valent to 83*234 of jicroxide of rhodium, which of course con- 

 tains 14 of oxygen. 'i'here remain of course 16"76"6 for the 

 water, wliich contains 11-8 of oxygen. Hence it would appear, 

 unless the preceding supposition be inaccurate, that in the hydrate 

 the oxide and water contain equal quantities of oxygen. If we 

 consider the jirotoxide as R!i + O, the volume of rhodium will 

 weigh 14'J0-3l.* 



{To be condnuc.il.) 



• As rLodium ii. a substance very liuU- kiinnii, and as few chemists will pro- 

 Imhiv jikv.' ail o|i|iorluiiity of .txainiiiiiig 11, I shall Ik re stale sninn observalioOi 

 whi< h I had urcnsiun lo make iluiiii;; rlif ('\|u-i iiiiiMils pivcii in tlic text. 



I I'll ill Ijcgiii ttithlho drsiilj.lli.ii ..f the oxiilis of rhodiuia. 1. I'rotuxidt. To 

 obtain tbi» oxide, reduce llif iiii-tul lo the ktatc of powdei-, and expose it in an 

 opcti vc>»cl tu a wudciutf nd heat, 'I'lic mital bcconus gruduallj blacit, and the 



