[ 277 ] 



by the fimple heat of a candle ; fome water alfo paffed, which un- 

 doubtedly abforbed fome air, a greater heat could not be applied 

 without Hiking a decompofition of the alkali itfelf; hence loo 

 grains of this fait would yield 80 of air, which in thefe circum- 

 ftahces would weigh 14,52 grains. In another experiment I ob- 

 tained fiill lefs of this air, for 50 grains of this fait afforded only 

 34,962 cubic inches, the barometer indeed flood higher, namely at 

 30,26, and the thermometer only at 53. 



Finding this method inadequate to the difcovery of the exadl 

 quantity of vol-alkali in this fait, I tried the effedt of fpontaneous 

 evaporation on a mixture of this fait with lime and water, but 

 foon found the quantity evaporated fo great that it was very evi- 

 dent it did not proceed from the mere yolatilization of the alkaline 

 part, but in a great meafure from that of the water alfo, hence I 

 was obliged to content myfelf with deteding the proportion of the 

 acid part. 



For this purpofe I made a folution of 400 grains cryflallized ni- 

 trous ammoniac, and to this added a fmall proportion of a folu- 

 tion of tartarin flightly aerated ; as the point of faturation could 

 not be afcertained by any tcft, I added but little of the tartarin, , 

 and fet the liquor to evaporate in a very gentle heat. The next 

 day I found fome cryflals of nitre, which I carefully picked out, 

 wafhed and dried, then added more tartarin to the mother liquor, 

 fet it to evaporate and cryflallize as before. Thus I proceeded for 



feveral # 



