1314,] Brfinite Proportions in Chemical Comlinatlons. 135 



and by the laws of definite proportions, I flatter myself that the 

 following table will furnish a much nearer approximation to the 

 truth than any hitherto offered to the public. 



The weight of an atom of azote whieh 1 chose in the Annals oj 

 Vhilosoplvj, vol. ii. p. 42, for the reasons tliere assigned, cannot, 1 

 find, be reconciled to the constitution of the nitrates. I have been 

 obliged, in consequence, to make the following alterations in the 

 weiglit of that atom, and in the weight of the bodies into which it 

 enters. 



iS'umber of Wiiglit of an 



jitoms. integrant particle. 



Azote 1-803 



Nitrous oxide 1 o + 1 « 2-803 



Nitrous gas 2 o + 1 a 3-803 



Nitrous acid 3 o + 1 a 4-S03 



Nitric acid .> o + 1 a fi-S03 



Ammonia ,. 1 h + 1 a 1 '^35 



According to these weights, nitrous gas must be a compound of 

 lOt) volumes of oxygen and 102-G volumes of azote, instead of 

 equal bulks. 



Genus II. — Nitrates. 



Number of Wciglit of an 



atoms. integrant particle, 



1201. Nitrate of potash 1 72 + I p 10-0S2 " 



202. Nitrate of soda 2 n + \ s 21-48S " 



203, Nitrate of ammonia .. \ n + I a 8'868 *= 



' Potash, when carefully dried, contains no water of crystalliza- 

 tion. I find that lUO grains of dry nitre, when decomposed by 

 sulphuric acid, give 83-6 grains of fused sulphate of potash, which 

 is equivalent to 45-6 per tent, of base. Hence nitre is composed 

 of 100 acid + 83-823 base. Dr. Wollaston's scale of eq\iivalents 

 gives us nitre composed of 100 acid + 86-7^4 base, wiiieh he 

 informed me was the result of a synthetic experiment. Now 

 6"-803 : G •: 100 : 88-1. This agrees nearly witli Dr. Wollaston's 

 result. 



'' 6-803 X 2 : 7-8S2 :: 100 : 57-92, Now Wenzel found this 

 fait composed of 100 acid + GO base, 1 have reason to believe 

 tiiat 100 acid + 57-8 base is very near the truth. Both of these 

 correspond sufficiently with the numbers in tlie table. How Ber- 

 tliol let obtained, as tlie result of analysis, 102'44 acid + 100 base, 

 1 cannot conceive. 



'- This supposes the weight of an integrant p:\rticlc of ammonia 

 to be l-yj5, and that it is composed of 1 /i + 1 «, Now G-803 : 

 I -93.5 :: 100 : 28-4; and Ikrzelius found tlic salt composed of 100 

 acid + .'51-2G6 base ((iilbcrt's Annalen, 1812, p. 1G5). Of prc- 

 U'ding analyses, Kirwan's ieenii- lu approach ncarcU the truth. 



