Carbonates of alumina, glucina, iron, chromium, etc. 329 



be obtained pure, may be with probability inferred from the 

 composition of its double salts with carbonate of potash and with 

 carbonate of ammonia, in which it exists, according to Ebelmen,* 

 as U^Og, CO^. On the other hand, as the salts of the sesqui- 

 oxyd of uranium, with the stronger acids, contain only one 

 equivalent of acid, instead of three like those of the other ses- 

 quioxyds, analogy with the carbonates just described should 

 make the carbonate of uranium contain less than one equivalent 

 of acid to one of base. The first view is, however, I think, the 

 more probable one. 



Summary of the above results. 



The following is a brief summary of the principal results 

 arrived at in the foregoing pages: 



1. All the carbonates in question, except, perhaps, the carbon- 

 ate of uranium, lose carbonic acid readily, during the processes 

 of washing and drying. To this may, in a great measure, be 

 attributed the discordant results of previous observers. 



2. The precipitate, produced by the alkaline carbonates in 

 solutions of the persalts of iron, has the composition FejO,, COj. 



3. The precipitate by the alkaline carbonates, in the violet 

 salts of the sesquioxyd of chromium, varies somewhat in com- 

 position, but approaches the formula CraO^, 200^ and is proba- 

 bly a mixture of this with a little of the more basic salt next 

 mentioned. The results of previous observers render it nearly- 

 certain that the above precipitate, by washing and drying, is 

 converted into a more stable salt of the composition Cr^Og, COj. 

 This latter is formed, when the precipitation takes place at the 

 '^oJhng point. The precipitate in the green solutions of chro- 

 mium consists of carbonate, not of oxyd as stated by Lefort, and 

 tas probably the same composition as in the violet salts. 



4. The precipitate by the fixed alkaline carbonates, in the salts 

 ^^ alumina varies in composition, but approaches the formula 

 ^5.03, CO, and is probably a mixture of this with a more basic 

 salt. Both in this and in the case of the carbonate of chromium, 

 ^^^ normal salt may be obtained nearly free froni basic com- 

 pounds, by pouring the alumina or chromium solution into that 

 ^^ the alkaline carbonate, instead of the reverse, so as to have 

 ^vvays an excess of the alkali. . , , . 



5. The precipitate by the alkaline carbonates, m the salts of 

 glucina, has the composition G3O3, CO3.. . 



6. The precipitate by the alkaline carbonates, m solutions 

 of the sesquioxyd of uranium, is undoubtedly a mixture of 

 l^ranate of the alkali and carbonate of uranium. The latter, if 

 ^^ could be obtained pure, would very probably have the compo- 



* Ann. Chim. et Phys., [3], t, 206-208. 



