242 PHILOSOPHICAL TRANSACTIONS. [ANNO 1782. 



from the evaporation of the watery part, and the remaining 45 contained alkali, 

 water, and acid, in the same proportion as the lOOgr. of Glauber's salt, per- 

 fectly dried, abovementioned; then these 45 contained 13.iggr. of vitriolic 

 acid, 2 1 .87 of fixed alkali, and 9.94 of water; consequently 100 gr. of crystal- 

 lized Glauber's salt contain 13.19 of vitriolic acid, 21.87 of alkali, and 64.94 of 

 water. 



I also saturated this alkali with the dephlogisticated nitrous acid, and found 

 that 100 gr. of the alkali took up 57 of the mere nitrous acid in the experiment 

 I most depended on; but this quantity varied in some experiments a ihw grains, 

 being sometimes 60, and sometimes 63 gr.; so that I conclude the proportion of 

 this acid, taken up by the alkali, is nearly the same as that of the vitriolic acid. 

 Supposing this quantity to be 57 gr. then 100 gr. cubic nitre, perfectly dry, 

 contain 30 of acid, 52.18 of alkali, and 17.82 of water; but cubic nitre crys- 

 tallized contains something more water; for 100 gr. of these crystals lose about 

 4 by gentle drying; therefore lOOgr. of the crystallized salt contain 28.8 of 

 acid, 50.09 of alkali, and 21.11 of water. 



Of mere marine acid, 100 gr. of this alkali required from 63 to 66 or 67 gr. ; 

 perhaps one reason of this variety is, that it is exceeding hard to hit the true 

 point of saturation. Allowing it to be 66 gr. then lOOgr. of perfectly dry 

 common salt contain nearly 35 of real acid, 53 of alkali, and 13 of water; but 

 lOOgr. of the crystallized salt lose 5 by evaporation; then lOOgr. of these 

 crystals contain 33.3 of acid, 50 of alkali, and 16.7 of water. 



The proportion of fixed air, alkali, and water, in crystallized mineral alkali, 

 I investigated thus: 200 gr. of these crystals were dissolved in 240 of water; 

 the solution was saturated by such a quantity of spirit of nitre as contained 40 

 of mere nitrous acid; hence I inferred, that these 200 gr. of alkali contained 70 

 of real alkali. The saturate solution weighed 40 gr. less than the sum of its 

 original weight, and that of the spirit of nitre added to it; therefore it lost 

 40 gr. of fixed air. The remainder therefore of the original weight of the 

 crystals, must have been water, that is, 90gr.; consequently lOOgr. of these 

 crystals contained 35 of alkali, 20 of fixed air, and 45 of water. This propor- 

 tion is, particularly with regard to the alkali, very different from that found by 

 Mr. Bergman and Lavoisier, which I impute to their having used soda recently 

 crystallized. Mine had been made some months, and probably lost much water 

 and fixed air by evaporation which altered the proportion of the whole. Accord- 

 ing to the calculation of these philosophers 100 gr. of this alkali takes up 80 of 

 fixed air. The specific gravity of the crystallized mineral alkali, weighed in 

 ether, I found to be 1 .42 1 . 



Of the volatile alkali. — It is not possible, by the old chemical methods, to 

 find the proportion of the ingredients in volatile alkalis, whether in a liquid or in 



