30 Experiments and Observations on 
sp. gr. 1.043. It must be observed that, this 
last is half equal in acid to my other tests, but 
has only } the power as an alkali, so that 400 
measures of it are required to saturate 100 of 
acid. 
Remarks on the subtricarbonate, subtetracar- 
bonate, subpentacarbonate, &c. of ammonia. 
The salts hitherto considered have been the 
carbonate of ammonia, that is, the compound 
of 1 atom of acid and 1 of base; the subcar- 
bonate, or perhaps more properly subbicarbo- 
nate, which is one atom of acid and 2 of base: 
it remains now to enquire whether the salts 
with 1 acid and 3, 4,5, &¢. base, have any 
existence. 
That the subtricarbonate exists 1 think 
there is little doubt. Dr. Priestley found 
that one measure of carbonic acid gas ab- 
sorbed nearly 3 measures of am. gas, in’ some 
cases; if itabsorb as much as 22, it would 
be sufficient to form a subtricarbonate. I 
find that 400 measures of subcarbonate solu- 
tion of 1.083 added to 170 measures of am- 
moniacal solution, .97, will form a solution 
containing the elements in due proportion to 
constitute a subtricarbonate; if no chemical 
union took place the sp. gr. of the mixture 
‘onght to be 1.049; but it was in fact 1.058 ; 
