124 HEINRICH ROSE ON THE COMBINATIONS OF 



quantity of water in the portion left behind in the retort in- 

 creases by its formation, so that sesquicarbonate and neutral 

 salt might be obtained in solution from it; for I have men- 

 tioned above, that the sesquicarbonate, with a greater quan- 

 tity of water, separates from the mother Hquor which contains 

 the neutral salt dissolved, which neutral salt is always formed 

 with the disengagement of carbonic acid gas, when the solution of 

 any combination of carbonic acid with ammonia, containing more 

 carbonic acid than the neutral salt, is heated for a long while. 

 The evolution of the carbonic gas during the slow sublimation of 

 the sesquicarbonate is then caused through the formation, not 

 only of the f carbonate, but also of the neutral salt. If, how- 

 ever, it be admitted that the sesquicarbonate is a double salt 

 of the carbonate and bicarbonate, it is then only the latter that 

 loses a portion of its carbonic acid at the renewed sublimation. 

 If the sublimation of the sesquicarbonate is further continued, 

 till what is contained in the retort has changed into a clear 

 liquid, the quantity of dissolved sesquicarbonate in it keeps de- 

 creasing, and at last the liquid consists of mere water, which 

 only contains a very little dissolved carbonate of ammonia, and 

 this is neutral. The sublimed salt in the neck and throat of the 

 retort gradually dissolves in the hot water evaporated. 



If this experiment be made with smaller quantities of common 

 sesquicarbonate, or with mixtures of sal-ammoniac and dry carbo- 

 nate of soda, heated in a small retort (the neck of which is con- 

 nected with a long glass tube) gradually, but continued until the 

 condensed water begins to show itself in the inclined glass tube, 

 then, after having become perfectly cool in the neck of the re- 

 tort, and in the part of the glass tube which is nearest to it, 

 and in which no solid salt has sublimed, crystalline needles of a 

 salt are deposited. I have often analysed this salt, but have 

 obtained very ditferent results in the several analyses. It seems 

 usually to be the sesquicarbonate combined with the greater 

 portion of water ; for I obtained from 0'1235 grm,, by a solution 

 of chloride of barium and ammonia, 0*247 grm. of carbonate of 

 bai-ytes, which corresponds to 44*83 per cent, carbonic acid; 

 0-284 grm. of another quantity gave 0-387 grm. metallic platina, 

 corresponding tj 23*70 per cent, ammonia. But sometimes it is 

 only the common sesquicarbonate, for in another experiment I 

 obtained 1-582 grm. of the sublimate,2-500 grm. of metallic platina, 

 or 27-48 per cent, of ammonia, (Here I must remark that the 



