410 On definite Pnportions. 



tity. This circumstance renders the analysis of soda still 

 more difficult, since the anialgaiii can scarcely be put into 

 the vessel in which it is to be weighed, without an increase 

 of its weight by the moisture wliich it acquires. Hence 

 the results of my experiments on soda agree still less with 

 each other than the foregoing, although they were per- 

 formed with equal accuracy, and founded on the same bases. 



1.) From 28 grammes of amalgam, by digestion with 

 water and a little muriatic acid, which docs not here, as in 

 the case of poiassium., produce a fetid hydrogen gas, I ob- 

 tained 'ISSS for the quantity of sodium. Tlie soda afforded 

 •365 gr. of fused common salt, which indicates '198 gr. of 

 dry soda^; whence we have for 100 parts of soda exactly 

 70 of sodium and 30 of oxygen. 



2.) From 37 gr. of amalgam I obtained '175 of sodium, 

 whence •■46 gr. of fused salt were forUiCd, containing •2496 

 of pure soda. According to this experiment, soda con- 

 tains 70" II of its base, and 29-89 of oxvgen. 



3.) Front 76 gr. of amalgam I obtained •439 of sodium, 

 which afforded l^llS gr. of fused salt ; containing '6066 

 of pure soda. Hence we have, for 100 parts of soda, 72^37 

 of sodium and 27^6;i of oxygen. 



The last experiment having been performed with the 

 largest quantity of the substance, and at the same time 

 with the gr' atest accuracy of which analyses of this kind 

 are susceptible, 1 think it probable that its result comes 

 nearest to the truth. Several other experiments, performed 

 with smaller (piaiitilies, gave proportions of oxygen varying 

 from 27 I" 36 per cent. It seems superfluous to describe 

 them more particularly, as they are in all respects less to 

 be depended on than this. 1 always found the proportion 

 of oxygen apoarently the greater as the quantity of the 

 base, with uhich I worked, was smaller. 



If we calculate the composition of soda in the same 

 manner as we have done that of potass, we shall find here 

 a similar agreement. 



Bucholz obtained, from 1000 grains of crystallized sul- 

 phate of si-ida, 698 01 sulphate of baryta, and he attributes 

 to the salt 568 grains 'f water of crystallization. Conse- 

 quently 100 parts of dry -sulphuric acid iTiust require for 

 their saturation 82'09 of dry soda. But several calcula- 

 tions for difllerent salts, proceeding with these numbers, 

 convinced me that there inust be some error in them, which 

 prohablv arises from the uncertainty respecting the dryness 

 of the Glauber's salt, and the quantity of its water of cry- 

 stallization, the latter not being capable of so accurate a 



determination 



