82 ALKALIMETRY 



tion to tho quantity of sulphate of soda present; an increase of density so much tho 

 more readily observable, that tho solubility of the sulphate of potash is greatly 

 augmented by tho presence of sulphate of soda. It had at first been thought that, in 

 order to obtain anything like accuracy, it would be necessary to combine all tho 

 potash with one same acid, preferably sulphuric acid ; and, consequently, that as the 

 potash of commerce always contains a little, and sometimes a rather considerable 

 quantity, of chloride of potassium, the latter salt should first be decomposed. Further 

 experiments, however, established tho fact that in dissolving chloride of potassium in a 

 saturated solution of sulphate of potash, tho specific gravity of tho liquor is not 

 materially increased, since the introduction of as much as 50 per cent, of chloride of 

 potassium does not increase that density more than 3 per cent, of soda would do when 

 examined by the natrometer a degree of accuracy quite sufficient for commercial 

 purposes. When soda is added to a saturated solution of sulphate of potash, tho 

 further addition of chloride of potassium thereto renders the specific gravity of tho 

 liquor less than it would have been without that addition an apparent anomaly duo 

 to the fact that chlorine, in presence of sulphuric acid, of potash, and of soda, combines 

 with the latter base to form chloride of sodium ; and it is this salt which increases the 

 solubility of sulphate of potash, though in a somewhat less degree than sulphate of 

 soda. Thus, if to a saturated solution of sulphate of potash 0'1-t of soda bo added 

 along with 0'20 of chloride of potassium, tho natromoter indicates only 0'125 of soda. 

 Seeing, therefore, that in such an exceptional case tho error does not amount .to more 

 than 0'015 of eiror, it will probably be found unnecessary in most cases to decompose 

 the chloride contained in the potashes of commerce, that quantity being too small 

 to materially affect the result. Yet, as the accurate determination of soda in potash 

 was a great desideratum, M. Pesier contrived two processes, one of which, in tho 

 hands of the practised chemist, is as perfect as, but much more rapid than, those ordinarily 

 resorted to ; the other, which is a simplification of the first, yields results of sufficient 

 accuracy for all commercial purposes. 



60. First process. Take 500 grains of a fair average sample of tho potash to bo 

 examined, dissolve them in as littlo water as possible, filter, and wash tho filter until 

 the washings are no longer alkaline. This filtering, however, may bo dispensed with 

 when the potash is of good quality and leaves but a small residue, or when an extreme 

 degree of accuracy is not required. 



61. The potash being thus dissolved, a slight excess of sulphuric acid is added 

 thereto ; the excess is necessary to decompose the chlorides and expel the muriatic 

 acid. The liquor so treated is then evaporated in a porcelain capsule, about six inches 

 in diameter ; and when it begins to thicken, it should be stirred with a glass rod, in 

 order to avoid projections. When dry, the fire must be urged until the residue fuses, 

 and it is then kept in a state of tranquil fusion for a few minutes. The capsule should 

 then be placed upon, and surrounded with, hot sand, and allowed to cool down slowly, 

 to prevent its cracking, which would happen without this precaution. 



62. The fused mass in tho capsule having become quite cold should now bo 

 treated with as little hot water as possible, that is to say, with less than 3,000 grains 

 of hot water ; and this is best done by treating it with successive portions of fresh 

 water. All the liquors thus successively obtained should then bo poured into a 

 flask capable of holding about 10,000 grains of water, and the excess of sulphuric 

 acid must be accurately neutralised by a concentrated solution of pure carbonate 

 of potash that is to say, until tho colour of litmus-paper is no longer affected 

 by the liquor, just as in ordinary alkalimetrical or acidimetrical assays. During this 

 operation, a pretty considerable precipitate of sulphate of potash is, of course, produced. 



63. The neutralising point being exactly hit, a saturated solution of sulphate of 

 potash is prepared, and brought to the atmospheric temperature ; a condition which 

 is expedited by plunging the vessel which contains the solution into a basin full of 

 cold water, and stirring it until the thermometer plunged in the liquor indicates that 

 the temperature of the latter is about tho same as, and preferably less than, that of 

 tho air, because in the latter case it may be quite correctly adjusted by grasping the 

 vessel with a warm hand. In order, however, to secure exactly tho proper tempera- 

 ture, the whole should be left at rest for a few minutes after having withdrawn the 

 vessel from the basin of cold water used for refrigerating it, taking care simply to 

 stir it from time to time, and to ascertain that tho thermometer remains at the samo 

 degree of temperature. This done, the liquor is filtered into a glass cylinder, c, on 

 which a scratch, H i, has been made, corresponding to 3,000 water-grains' measure. 

 If the directions given have been exactly followed, it will be found that the filtrate is 

 not sufficient to fill it up to that mark ; the necessary volume, however, should be 

 completed by washing tho deposit of sulphate of potash in the filter, B, with a satu- 

 rated solution of the same salt (sulphate of potash) previously prepared. It is advis- 

 able to use a saturated solution of sulphate of potash which has been kept for some 



