Hajy Method of rendering Water perfe^ly pure. 113 



dite- water of wells in places where rdenite or plaifier abounds, but which are neverthelefs 

 fir from being pure, and are neceffarily fubject to vary according to the quantity of raia 

 water whiciv dilutes that which has remained upon the foJuble matter. 



Recouffe is therefore bad to diftillation, to purify the water employed in laboratories ; but 

 if on the one hand we confider the labour and expence it rt-quires, and on the other band the 

 quantity which it is lieceilary to have in readinefs for the fmalleft operations; it will not 

 appear furprifing, when I affert, that there are few days in which a chemift does not avoid 

 making fomc experiment, or obtain uncertain refults, for want of having this article in his 

 ppwer. ^ It is only-wi^h diftilled water that perfeft reagents can be prepared ; diftilled water 

 is confumcdirt iiitufions, macerations, foluiions, and edulcorations ; repeated lotions demand 

 a- large quantity of this fluid ; it ought to be ufed even for rinfing veffels ; and in order to 

 avoid deceitful conclufions, it is even proper to ufe it in the hydropneumatic veflels. 



I have long been in the habit of fupplying moft of thefe demands with rain water ; not • 

 •with that which is diredly received, though in faft it is of confiderable purity, particularly 

 in countries where there is no reafon to fear that plaifter fliould rife with the dud ; but the 

 quantity of this would be too fmall. I have, therefore, ufed water colle£led with care from ^ 

 the roofs of houfes after the rain has wafhed the furface. 1 filter it without delay ; and in 

 this manner from time to time I obtain a confiderable provifion, without much labour and at 

 no expence. But it is evident, that in order to depend on the purity of' this water, it is^ 

 neceffary that there fliould be no gypfeous matter in the compofition of the mortar of the 

 roof, or the plaiftering of the chimneys. This condition does not obtain at Paris ; info-- 

 much that I have fometimes found the firft water of the "gutters more felenitous than that 

 of the Seine in the time of floods. I have thought of a method of fuppljing this refource, . 

 by a procefs which may be ufed in all countries ; and the fuccefs which has attended my 

 trials renders it a duty to communicate it, in order to place the inftruments of analyfis in > 

 the hands of a greater number of operators. 



Rain water collected from the roofs of houfes which hive been prevsoufly wafhed, cannot^' , 

 and in fa£i: does not, contain any thing but the very fmall portion of fulphate of lime which • 

 ithas taken up during its contaft with the plaifter of the chimneys and the pointing of the 

 ridges.' It is neceffary, therefore, to deprive it of this, in order to have water in a very pure - 

 ftate. For this purpofe I prepare a folution of barytes according to the procefs * by which 

 our colleague Vauquelin has rendered fo eafy what Bergman attempted with imperfeft fuc- - 

 cefs. I pour this folution into the filtered rain water, until, after the precipitation has fub- 

 fided, the laft drop exhibits no alteration of tranfparency. I even add a^mall quantity in ■■ 

 excefs, which fliews itfelf by the vinous colour it gives to paper coloured with fernambouc 

 or brazil wood. Thisexcefsfoon falls ;down in the ftate of carbonate of barytes, by fimple ' 

 expofure to the ain The precipitation may be very fuddenly determined, by the addition of : 

 water impregnated with carbonic acid. Too much muft not however be added, becaufe it-: 

 would take up a portion of the precipitate. But the fpontaneous evaporation of -the excefs ^ 

 of gafeous acid in the open air would in this cafe foon reftore its purity.. . 



To judge with whatfacilityand.trifling expence the whole of the pijrc water fbtcxperi-;- 



*i 



'Philofophical Journal, I. 53;. Dr. H6pe in the Edinburgh Tranfailidns, iv. 36; informs us, that, con- • 

 trary to the alTertions of many chcmifts, the native carbonate of barytes may be deprived of its acid by mere- 

 Aeat (,of a froith's forge, in a black lead crucible), N, . 



