<0 Mr. kirwan’s Experiments and Observations on 
part of the former acids being expelled by the heat of evapora- 
tion, the neutral lalts begin to cryftallize, and confeqnently 
give out heat ; but the vitriolic being then in greater proportion 
re-acts on thefe falts, reftores their fpecific fire to their acid 
principle, and recombines with their alkaline balls, as already 
explained. 
Hence, though allum is in reality decompofed by the nitrous 
and marine acids, yet when the folution of it in either of thefe 
acids is evaporated to a certain degree, the vitriolic acid, of 
which it contains a larger proportion than any other terrene 
fait, re-adts on the nitrous and marine allums, and expels their 
acids, as Mr. chaptal has Ihewn. 
In explaining thefe phenomena I have all along fuppofed the 
dodhlne of Dr. black to be well known, viz. that lolids ab- 
forb heat during their folution. Both the heat and cold, pro- 
duced in different folutions, feem to me to depend on the fame 
principle. If the menftruum gives out only fo much of its fire 
as the lolvend can abforb, or lefs, then cold is produced ; but if 
it gives out more of its fpecific fire than the folvend can abforb, 
this furplus becomes fenfible, and affedfs the thermometer by 
producing heat in proportion to its quantity. 
Of the affinity of the mineral acids to metallic fubjiances. 
Having thus, in every inftance, eftablifhed the agreement 
betwixt the quantity of any alkaline or terrene bails, taken up 
at the point of faturation by a given weight of any of the 
three mineral acids, and the quantity of affinity which each of 
thefe acids bears to fuch balls, I naturally extended my views 
to metallic fubftances, to try whether this coincidence could be 
traced with regard to them alfo ; but the difficulties that oc- 
curred in this inquiry were fo great, that the fame degree of 
certainty muff not be expedted as in the foregoing part. 
