QUANTITY OF WATER IN FUSED SODA. 353 



quantity. Nitrate of siiyer is such a powerful test of 

 muriatic acid, that it detects extremely small quantities. 

 Kirwan found, that one part of muriatic acid diluted with 

 108333 of water could be detected by nitrate of silver*. 

 Does muriate of lime indicate such imperceptible quan- 

 tities of carbonic acid, particularly in a liquid that is not 

 neutral ^ 



These observations seem to me to answer entirely the ob- 

 jections Mr. d'Arcet has made to my analysis : though at 

 the same time I am far from considering his method as a bad 

 one. Whence then, perhaps, it will be asked, arises the 

 difference between the results? I conceive it originates <^'^"^^* °f ^^'^' 



o 1 • i.inxjA 1 -, »/. ferencfe iii the 



from the proportions,, which Mr. d'Arcet has adopted for results., 

 the carbonate of lime, and from the difficulty he must have 

 found to deprive this salt entirely of water, without ex* 

 pelling some of its acid. 



Thenard and Biot have just made a comparative analysis Component 

 of carbonate of lime and arragonite. In their experiments |^^(g ^f I'me, 

 they employed all the care, that might be expected from 

 such experienced chemists. From these it appeared, that 

 5G of lime unite with 43 of carbonic acid. 



But IMr. d'Arcet has found, that 100 parts of crystallized Calculation 

 subcarbonate of soda, containing 36*39 of dry subcar- 

 bonate, gave 34*81 of carbonate of lime, which, exposed 

 to a strong fire, left 18 72 of quicklime. This quantity 

 of lime, from the analysis above quoted, would combine 

 M'ith 14'37 of carbonic acid, which were accordingly con., 

 taincd in the 3C'39 of dry subcarbonate : and my analysis 

 would have given 13-63. If it be considered too, that I 

 was obliged to take that analysis of IMr. d'Arcet's, in which 

 he found most carbonate of lime, these results, I imagine, 

 will not be thought very wide. 



iVow if Mr. d'Arcet's experiments be calculated from my hrin*'; the re- 

 analysis of the subcarbonate of soda, it will appear, that^"'^* ^'"^ "'^^'^* 

 100 parts of the soda he analysed contain 20 of waterj 

 while I found 18'86. 



* r am aware, that Mr. Borlbollet junr. has detected much 

 smaller quantities of acid with the yarne test. It is true, he found, 

 that muriate of silver was soluble in concentrated and boiling solu- 

 tions of almost all muriates : but it is sufficient to dilute iheni with 

 water, to occasion its reappearance. 



Vol. XXVII.—Suj-pi-kment. A a IV. On 



