86 ALKALIMETRY 



( 76) into the glass beaker containing the aqueous solution of the soda ash under 

 examination, until complete neutralisation is attained, stirring briskly all the time, or 

 after each addition of the test acid. The neutralisation point being hit, the sides of 

 the alkalimeter are allowed to drain, and the operator then reads off the number of 

 divisions employed, which number indicates the per-centage of real soda contained in 

 the sample assayed. Thus, if the sample operated upon be the same as that alluded 

 to before, the number of divisions employed being 52 would indicate 52 per cent, of 

 real soda, 



81. If the operator wishes to estimate the amount of soda in the sample as car- 

 bonate of soda, he should fill the alkalimeter with the test acid in question (specific 

 gravity 1-268) up to the point marked carbonate of soda, and fill the remaining 

 divisions with water, shake the whole well, and proceed with the neutralisation of the 

 aqueous solution of the sample in the glass beaker as just described. Supposing, as 

 before, that the sample in question contains 52 per cent, of real soda, it will now be 

 found that the number of divisions employed altogether to saturate the sample com- 

 pletely are very nearly 89, for 52 of caustic soda correspond to 88'90 of the carbonate 

 of that alkali. 



82. If the soda ash is very poor, instead of operating upon 1,000 water-grains' 

 measure, or one-tenth part of the whole solution ( = 100 grains' weight of the soda 

 ash, 76-77), it is advisable to take three or four thousand water-grains' measure of 

 the alkaline solution, and to divide, by three or four, the result obtained by saturation. 

 Suppose, for example, that the quantity of real soda found is 46 ; this, if only 1,000 

 grains' measure had been taken, would, of course, indicate 46 per cent. ; but as 4,000 

 water-grains' measure of solution have been taken instead, that number 46 must, ac- 

 cordingly, be divided by 4, which gives 11^ per cent, only of real soda contained in 

 the sample under examination. 



83. The soda ash of commerce contains generally a per-centage of insoluble sub- 

 stances, which are removed by filtering, as wo said, and a greater or less quantity of 

 chloride of sodium (common salt) and of sulphate of soda, which, however, do not in 

 the slightest degree interfere with the accuracy of the result. But there is a source 

 of error resulting from the presence in the soda ash of sulphide of calcium, of sulphite, 

 and sometimes also, though more rarely, of hyposulphite, of soda. When sulphuret of 

 calcium is present in the ash, on heating the latter by hot water, a double decompo- 

 sition takes place : the sulphuret of calcium, reacting upon the carbonate of soda, forms 

 sulphuret of sodium and carbonate of lime. Now sulphuret of sodium saturates the 

 test acid just as carbonate of soda ; but as it has no commercial value, it is clear that 

 if the ash contains a quantity of the useless sulphuret at all considerable, a very serious 

 damage may be sustained by the purchaser if the per-centage of that substance present 

 in the ash bo taken account of as being soda. Sulphite of soda is produced from the 

 oxidation of this sulphuret of sodium, and is objectionable inasmuch that when the 

 test acid is added slowly to the aqueous solution of the ash, the effect is to convert 

 the sulphite into bisulphite of soda, before any evolution of sulphurous acid, and 

 consequently before the pink reaction on litmus-paper is produced. 



84. In order to obviate the inaccuracies resulting from the neutralisation of a portion 

 of the test acid by these substances, it is necessary to convert them into sulphates of 

 soda, which is easily done by calcining a quantity of the sample with five or six 

 per cent, of chlorate of potash, as recommended by Gay-Lussac and Welter. The 

 operator, therefore, should intimately mix 60 or 60 grains' weight of pulverised 

 chlorate of potash with 1,000 grains of the pulverised sample, and fuse the mixture in 

 a platinum crucible, for which purpose a blowpipe gas-furnace will be found ex- 

 ceedingly convenient. The fused mass should be washed, and the filtrate being 

 received into a 10,000 water-grains' measure, and made up with water to occupy that 

 bulk, may then be assayed in every respecfc as described before with one or other 

 of the test acids mentioned. 



85. When, however, the soda ash contains some hyposulphite of soda which for- 

 tunately is seldom the case, for this salt is very difficultly produced in presence of a 

 very largo excess of alkali it should not be calcined with chlorate of potash, because 

 in that case one equivalent of hyposulphite becomes transformed, not into one equivalent 

 of sulphate, but, reacting upon one equivalent of carbonate of soda, expels its carbonic 

 acid, and forms with the soda of the decomposed carbonate a second equivalent of sul- 

 phate of soda, each equivalent of hyposulphite becoming thus converted into two 

 equivalents of sulphate, and therefore creating an error proportionate to the quantity 

 of the hyposulphite present, each equivalent of which would thus destroy one equi- 

 valent of real and available alkali, and thus render the estimation of the sample 

 inaccurate, and possibly to a very considerable extent. 



86. When this is the case, it is therefore advisable, according to Messrs. Fordos and 

 Gclis, to change the condition of tho sulphurets, sulphites, and hyposulphites, by add- 



