of the fixed Alkalis . 1 07 



inches*, would become 186'67 measures, the weight of 

 which would be about -Ql Si grains troy f, but *0 184 : '1364:: 

 13*29 : 100; and according to this estimation 100 parts of 

 potash will consist of 86*7 basis, and 13*3 oxvgen nearly. 



In the second experiment *07 grains of the basis absorbed 

 at temperature 63? of Fahrenheit, and under pressure equal 

 to 30*1 barometer inches_, a quantity of oxygen equal in 

 volume to 121 grain measures of mercury, and the proper 

 corrections being made as in the former case, this gas would 

 weigh '01189 grains. 



But as -07 + '01189 = -08189:07 : : 100:85-48 nearly, 

 and 100 parts of potash will consist of 85-5 of basis and 

 14" 5 of oxygen nearly. And the mean of the two experi- 

 ments will be 86-1 of basis to 13*9 of oxygen for 100 parts. 



In the most accurate experiment that I made on the com- 

 bustion of the basis of soda -08 parts of the basis absorbed 

 a quantity of oxygen equal to 206 grain measures of mer- 

 cury ; the thermometer being at 56° Fahrenheit ; and the 

 barometer at 29 - 4 ; and this quantity, the corrections being 

 made as before for the mean temperature and pressure, 

 equals about -02 grains of oxygen. 



And as '08 -f -02 = -10 : -08, : : 100 : SO, and 100 parts 

 o£«soda according to this estimation will consist of 80 basis 

 to 20 of oxygen. 



In all cases of slow combustion, in which the alkalis were 

 not carried out of the tray, I found a considerable increase 

 of weight; but as it was impossible to weigh them except in 

 the atmosphere, the moisture attracted rendered the "results 

 doubtful; and the proportions from the weight of the oxy- 

 gen absorbed are more to be depended on. In the experi- 

 ments in which the processes of weighing were most speedi- 



* In the correction for temperature, the estimations of Dalton and Gay 

 Lussac are taken, which make gases expand about 7 j 6 of the primitive - vo- 

 lume for every degree of Fahrenheit. 



f From experiments that I made in 1799, on the specific gravity of oxygea 

 gas, it would appear that its weight is to that of water as 1 to 748, and ta 

 that of quicksilver as 1 to 10142. Researches Chrm. and'Phil. p. S; and wirfr 

 this estimation, that deducibie from the late accurate researches of Messrs, 

 Allen and I'epys on the Combustion of the Diamond almost precisely agrees, 

 phi!. Trans. 1307, p. L'75, 



