[ 96 j 
Experiment IT. 
1 1 28 ounces of Rathbcne-place water were (Mil- 
led in the lame manner as the former. The diflilled 
water was divided into two parcels, that parcel which 
came over firft weighing 12 1 ounces, the other 146. 
A preparatory experiment was firft made, in order to 
form a judgement of the comparative ftrength of 
each parcel, and alio of the quantity of acid which it 
would require to faturate them. This was done by 
dropping l'ugar of lead into each parcel till it ceafedto 
make a precipitate. It was judged from hence that 
the firft parcel contained about 2.4 times as much 
volatile alcali as an equal quantity of the fecond. Into 
30 ounces of the firft parcel, mixed with as much of 
the fecond, was then put 43 grains of oil of 
vitriol, which was fuppofed to be about 4 more than 
fufticient to faturate the alcali therein. The mixture 
was then evaporated. When reduced to a fmali 
quantity, it was found to be rather acid : fixteen 
grains of volatile fal ammoniac were therefore added, 
which feemed nearly fufticient to neutralize it. Being 
then evaporated to drynefs, it left fixty-fix grains of a 
brownilh fait, which diftolved readily in water, 
leaving only a trifling quantity of brown fediment. 
A little of this fait was found to make no precipi- 
tate on the addition of fixed alcali, and the remainder, 
being boiled with lime, was converted into felenite ; a 
fure fign that the fait was merely vitriolic ammoniacal 
fait. The volatile alcaline fait contained in fixty-fix 
grains of vitriolic ammoniacal fait is 584. grains ; from 
whence deducing fixteen grains, the weight of the 
volatile fal ammoniac added, it appears that the diflilled 
water 
