etl a ancien et 
WATERS from ICELAND. his g 
water, does not precipitate filver * ; and to prevent any part of 
the filver being precipitated by the alkali of the water, I added 
of purified aquafortis, more than enough to faturate the alkali, 
before I added the folution of filver. 
ANOTHER fet of experiments, on the fame plan, but made 
with Glauber’s falt and the folution of barytes, in place of com- 
mon falt and folution of filver, enabled me to afcertain with equal 
exatnefs the quantity of Glauber’s falt contained in thefe waters. 
I firft learned that if pureGlauber’s falt be perfectly exficcated, by 
evaporating the water that is in its cryftals, 10 parts of this 
exficcated falt are fufficient to precipitate as much barytes, from 
its folution in muriatic acid, as will form 17 of barytes vitrio- 
lica. This fact being afcertained, I added fome of the diffolved 
barytes, to feparate portions of the Iceland waters, fo long as 
any muddinefs and precipitation was produced ; and I care- 
fully collected, wafhed, dried and weighed the precipitates. 
I thus learned, that the water of Rykum contains in gr. 10,000 
of it, as much Glauber’s falt as would give gr. 1.28 of exficcated 
Glauber’s falt, and the water of Geyzer as much as would give 
gr. 1.46 f. 
‘In making thefe laft experiments alfo, I added fome purified 
nitrical acid to the Iceland waters, to prevent any precipitation 
of the barytes which might have been occafioned by the alkali. 
of the water. 
In reviewing the experiments I have now defcribed, if we. 
neglect the fmall quantity of fulphureous gas, the contents of ° 
thefe waters will appear as follows : 
In: 
* See the appendix to this paper.. 
+ The method by which thefe {mall quantities of fediments and precipitates were col- 
leGted and weighed, is explained in the appendix to this paper. 
