322 Cai-honates of alumina, glucina, iron, chromium, etc. 



dry air and the water caught by a weighed chlorid of calcium 

 tube. The loss of weight in the bulbed tube and substance, mi- 

 nus the increase of weight in the chlorid of calcium tube, would 

 give the weight of carbonic acid corresponding to the carbonate 

 of iron, while the carbonate of soda and sulphate of soda would 

 remain unchanged. The sesquioxyd of iron, remaining in the 

 bulb, was washed and weighed. Of course, by this method, the 

 water cannot be estimated, which, according to all previous ob- 

 servers, exists in all the carbonates in question. In my first 

 analyses the substances, though not washed, were dried over 

 sulphuric acid : but three analyses of the same substance, wliieli 

 gave successively the percentages of carbonic acid, IS-Ti, li'tt') 

 and 12-95, showed conclusively that the substance lost carbonic 

 acid by standing. All the other analyses made, both of this and 

 of the other carbonates, were therefore made with the precipi- 

 tate, while still moist. The formulas obtained, therefore, express 

 simply the relation between the acid and the base, and a certain 

 amount of water must probably in every case be understood. 

 In every case, more than one preparation was analyzed, to ascer- 

 tain whether the precipitate had always the same composition. 



Of the carbonate of iron four analyses were made, with the 

 following results : 



re^Oa 80 78-43 7988 79-47 79-09 80-52 



CO, _22_ _2r57_ _20-12 20-S jo^ ^I'l 



102 100-00 lOO-Oo' lOO-Oo" 10000 10000 



Considering the ease with which the substance loses carbonic 

 acid, the above results agree with the calculated percentage prob- 

 ably as closely as could be expected— closely enough, at any 

 rate, to fix the composition of the substance with sufficient cer- 



2. Carbonates of the sesquioxyd of chromium. 

 Cr^Os, 'iCOo 

 CraOa, CO. 



The earliest analysis of carbonate of chromium is by Meiss. 

 who found it to contain 77-30 Ct,0„ 15-54 CO^ and T-l'- 

 Meissner's analysis, however, is quite untrustworthy, as un- 

 seen from his details of it. According to Lefortf the prec ■ 

 by carbonated alkalies in the violet salts of chromium is (' 

 C0,-f4H0; in the green modification only the hydrated ■ 

 According to Langlois,+ the precipitate by carbonated alka! ■ 

 ATTA ^^ chromium, dried at 100° C, is Ct^O^, C03+<-'- 

 6H0. Wallace! precipitates sesquichlorid of chromium bj- 



^ Chem. Gaz., 1858, 410, in Jahresb,"i858 70. 



