, 34 Mt~ KIRWAn’s Experiments, Sec. on the Specific Gravities 
maximum when the accrued denfity is at its minimum , but not 
reciprocally ; and hence the point of faturation is probably the 
maximum of deniity and the minimum of fenfible attra&ion of 
one of the parts. Hence no decompofition operated by means 
of a fubftance that has a greater affinity with one part of a com- 
pound than with the other, and than thefe parts have to each 
other, can be complete, unlefs the minimum affinity of this 
third fubftance be greater than the maximum affinity of the 
parts already united. Hence few decompofitions are complete 
without a double affinity intervenes ; and hence the laft portion 
of the feparated fubftance adheres fo obftinately to that to 
which it was firft united, as all chemifts have obferved. Thus, 
though acids have a greater affinity to phlogifton than the earths 
of the different metals have to it, yet they can never totally 
dephlogifticate thefe earths but only to a certain degree ; fo 
though atmofpheric air, and particularly dephlogifticated air, 
attracts phlogifton more ftrongly than the nitrous acid does ; yet 
not even dephlogifticated air can deprive the nitrous acid totally 
of its phlogifton, as is evident from the red colour of the nitrous 
acid when nitrous air and dephlogifticated air are mixed toge- 
ther. Hence alfo mercury precipitated from its folution in any 
acid, even by fixed alkalies, conftantly retains a portion of the 
acid to which it was originally united, as Mr. bayen has 
fhewn ; fo alfo does the earth of allum, when precipitated in 
the fame manner from its folution ; and thus feveral anomalous 
decompofitions may be explained. Indeed, I have reafon to 
doubt, whether mercury does not attradl acids more ftrongly 
than alkalies attradl them. 
4thly. That concentrated acids are, in fome meafure, phlo- 
gifticated, and evaporate by union with fixed alkalies. 
