HYDRATES AND CARBONATES OF THE ALKALI-METALS, ETC. 455 



potassium which the original salt in all probability contained, could be expected to 

 remain in the mother-liquor, with the rest of the impurities. The oxalate of soda was 

 collected on a filter, washed cautiously with water, dried, and ignited in platinum. The 

 residual carbonate was dissolved (in a platinum basin) in water, a trace of insoluble 

 matter filtered off, the filtrate evaporated to dryness in platinum, and the residue 

 dehydrated at a dull red heat without fusion, A salt prepared in this manner was used 

 for the atomic-weight determinations reported on at the end of last section. 



Experiment I. — 8*462 grm. of (Erfurt) salt was heated in a current of pure and 

 dry hydrogen for three hours by means of a powerful "Bunsen," which raised the 

 temperature of the salt to considerably above its fusing-point; and then for other two 

 hours over a gas blowpipe. The experiment was then stopped with the view of deter- 

 mining the weight of the residue; but as the salt was found to have crept very perceptibly, 

 besides having partially suffered volatilisation, the loss of weight observed is not worth 

 stating. After this interruption the heating was resumed (the Bunsen and the blowpipe 

 being used alternately), and continued until the crucible had been over the fire for in 

 all ten hours, including five hours of blowpipe work. Of the product 2'0171 grm. were 

 dissolved in water and the solution was made up to 18379 grm. This solution was 

 preserved in a bottle under a vaselined stopper and parts of it used for the following 

 analyses : — 



(1) As a preliminary test for caustic alkali, 50'434 grm. of the solution were mixed 

 with excess of chloride of barium, the mixture allowed to settle in the absence of air, and 

 the clear part examined. It was strongly alkaline ; an aliquot weight was titrated with 

 fifth-normal hydrochloric acid, which showed that the substance contained 13 '5 9 per cent, 

 of Na 2 0, uncombined with carbonic acid. 



(2) 50*399 grm. of solution, when decomposed with hydrochloric acid, &c, gave 

 0-1871 grm. of C0 2 and 07857 of sulphate of soda. 



(3) 50-396 grm. of solution gave 0'1892 of C0 2 and 0-7850 of Na 2 S0 4 . Hence we 

 have — 



(2) (3) Mean. Calculated* 



Carbonic acid, . . 33-83 34-20 34'02 34-07 



Soda, Na 2 0, . . 62-06 62'00 62-03 61-92 



Water, by difference, 4-11 3-80 3-95 4-01 



100-00 100-00 100-00 100-00 



These analyses consumed a considerable part of the product. The rest was put back 

 into the crucible, and again heated in hydrogen over the blowpipe for five hours, with a 

 view to seeing whether there would be any further diminution in the proportion of car- 

 bonate. Unfortunately, however, when the crucible came to be opened, it was almost 

 empty, and only 9 6 "2 mgrm. of the product could be scraped together for an analysis, 



* On the assumption that the substance is a mixture of 82 -16 per cent, of carbonate of soda, and 17 -84 per cent, of 

 hydrate NaOH, equal (the latter) to 13 - 83 per cent, of Na 2 0, which agrees fairly with the determination by chloride of 

 barium. 



