3 ^ An Attempt to make a 'Thermometer 
nor diffolves eafily in water, and which therefore would have 
cryftallized long before the alkali became dry, or remained 
after its deliquiation. 
A twentieth part of gypfum, ground with clay, was very 
diftinguifhable by both the foregoing procefles ; producing a 
lulphureous fmell, and calcareous earth by calcination with 
charcoal powdei ; and cryftals of vitnolated tartar by calcina- 
tion with the fame alkaline fait. 
To feparate the pure argillaceous part, or that matter which; 
in all clays forms alum with the vitriolic acid, 240 grains of 
this clay were thoroughly moiftened with oil of vitriol, boiled 
to dr\ nefs, and at laft made nearly red-hot. The mixture was 
then boiled in water ; the earth which remained undiftblved 
Avas treated again in the fame manner with vitriolic acid, and 
this opeiation repeated five or fix times. The clay was dimi- 
nifhed in the firf f operation about 70 grains ; but dels and lefs in 
the fucceeding ones, and in the laft fcarcely two grains. The 
filtered liquors yielded cryftals of true alum ; but its quantity 
was not examined, as the produce of alum from aluminous 
earth is already fufficiently known, and the quantity of alu- 
minous earth itfelf, or its proportion to the indiffoluble earth, 
was here the objech From the 240 grains of clay there re- 
mained in one experiment 98, and in another 95 grains of 
indiffoluble earth ; fo that five parts of this clay confift of 
three parts of pure argillaceous or alum earth, and two parts 
of an earth of a different kind. 
With refpeft to the nature of this laft earth, it is eafier to 
deteimine negatively what it is not, than pofitively what it is ; 
but afcei taming the former will be a great ftep towards the 
difcovery of the latter. 
2 
1 
That 
