VOL. LXXII.] PHILOSOPHICAL TRANSACTIONS. 243 



a concrete state; seeing that, though it may be separated from fixed air, yet it 

 cannot from water, on account of its extreme volatility. Then to find this pro- 

 portion we must recur to the experiments of Dr. Priestley, who by his new 

 analysis produced this alkali free from the aerial acid and water in the form of 

 air; and in the 3d volume of his observations, p. 2Q4, informs us, that 1-^ 

 measures of alkaline air take up, and are saturated by, 1 measure of fixed air.* 

 Let us suppose the measure to contain 100 cubic inches; then 185 cubic inches 

 of alkaline air take up 100 of fixed air; but 185 cubic inches of alkaline air 

 weigh, at a medium, 42.55 gr.; and 100 cubic inches of fixed air weigh 57 gr. ; 

 then lOOgr. of pure volatile alkali, free from water, take up 134 of fixed air. 



On expelling its aerial acid from a parcel of this alkali in a concrete state, and 

 formed by sublimation, I found 100 gr. of it to contain 53 of fixed air, and 

 therefore, according to the preceding reasoning, 39-47 of real alkali and 7.53 of 

 water per cent. Saturating a solution of this alkali with the vitriolic, nitrous, 

 and marine acids, I found, that lOOgr. of the mere alkali take up 106 of mere 

 vitriolic acid, 115 of the nitrous, and 30 of the marine. The specific gravity 

 of the concrete volatile alkali weighed in ether was 1.4076. 



The proportion of water in the different ammoniacal salts I have not been able 

 to find, on account of their volatility; but believe it to be very small, as volatile 

 alkali and fixed air crystallize without the help of water, when both are in an 

 aerial state. 



Of calcareous earth. — I first dissolved this earth in the nitrous acid, and found 

 that, after allowing for the loss of fixed air and the quantity of water I formerly 

 mentioned, 100 gr. of the pure earth take up 104 of mere nitrous acid. Instead 

 of dissolving this earth immediately in the vitriolic acid, I precipitated its solu- 

 tion in the nitrous by the gradual addition of the vitriolic, and found that to 

 effect this, Q\ or 92 gr. only of mere vitriolic acid were required. lOOgr. 

 of this pure earth demand for their solution 112 of mere marine acid. The 

 solution, which is at first colourless, grows greenish on standing. Natural 

 gypsum varies in its proportion of acid, earth, and water, 100 gr. of it con- 

 taining from 32 to 34 of acid, and also of earth, and from 26 to 32 of water. 

 The artificial contains 32 of earth, 29.44 of acid, and 38.56 of water; when 

 well dried it loses about 24 of water, and therefore contains 42 of earth, 39 of 

 acid, and 19 of water percent. 100 gr. nitrous selenite, carefully dried, con- 

 tain 33.28 of acid, 32 of earth, and 34.72 of water. 100 gr. marine selenite, 



* I have lately repeated this experiment, and found that one measure of alkaline air is saturated 

 by less than half of one measure of fixed air, but more than | ; conformably to Dr. Priestley's first 

 experiment, p. 293 ; by which it appears, that 100 gr. of alkaline air require about \20 of fixed au- 

 to saturate them : and hence 100 gr. of concrete volatile alkali contain about 53 of fixed air, 44 of 

 mere volatile alkali, and 3 of water. — Orig. 



1 1 1 



