Another method sometimes used is that of M. Mohr, and practised as follows:—The alkaline solution, slightly coloured blue with litmus, is strongly super-saturated with a standard acid (sulphuric or oxalic) of known strength, supplied from an alkalimeter in the usual manner; the last traces of carbonic anhydride being removed by boiling, shaking, blowing into the flask, and, finally, sucking
out the air. A standard solution of caustic soda (of a strength exactly corresponding to that of the test-acid already used) is now cautiously added, drop by drop, until the colour, rendered yellowish-red by the acid, just appears of a light blue. The difference between the quantity of the solution of the test-alkali and of the test-acid consumed, expresses the exact quantity of acid neutralised by the alkali, and hence also its strength.
Besides the above methods, the alkaline carbonates are analysed, by the loss of carbonic anhydride (carbonic acid) they suffer, by being decomposed by a strong acid. The best method in use is that of MM. Fresenius and Will, and depends on the same principle, and is performed in a similar manner and in a similar apparatus to that described under Acidimetry; the only difference being that here the uses of the small tube (e) is dispensed with, and that the alkali is tested under the form of carbonate, instead of bicarbonate.
Oper. The smaller flask (B) is about half filled with concentrated sulphuric acid, and the sample of alkali, in solution (under the form of carbonate), being placed in the larger flask (A), water is added until it is about one third full. The tubes are then fitted into the apparatus quite air-tight; the end of the tube (b) is fastened with a piece of wax, and the whole is very carefully weighed. The apparatus is now removed from the scales, and a perforated cork, or a small piece of india-rubber tube, being temporarily applied to the end of the tube (h), a few bubbles of air are sucked out of the flask (B) by means of the lips; the consequence of which is, that on removing the mouth the acid in (B) ascends to a certain height in the tube (c). If in a short time this little column of liquid maintains its height in the tube, it is a proof that the apparatus is perfectly air-tight, and as it should be. Suction is now again cautiously applied to the tube (h) and a little of the acid in (B) made to flow over into the flask (A), the quantity being proportionate to the vacuum produced by suction, and capable of being regulated at will. No sooner does the acid come into contact with the carbonate in the flask (A) than the evolution of carbonic acid commences, and this, from the construction of the apparatus, having to pass through the concentrated sulphuric acid, is rendered quite dry before it can escape by the tube (d) into the atmosphere. Whenever the effervescence flags, a little more acid is sucked over, until the whole of the carbonate is decomposed; after which an additional quantity is made to pass into (A), so as to raise the temperature considerably, for the purpose of expelling all the gas absorbed by the fluid during the operation. As soon as this is effected, the wax is removed from the aperture (b), and suction applied to (h), until all the carbonic acid in the apparatus is replaced by atmospheric air. The whole is now allowed to cool, and (together with the piece of wax removed) is again accurately weighed. The loss of weight gives the exact amount of dry carbonic anhydride, or anhydrous carbonic acid, which was contained in the specimen, from which the weight of PURE ALKALI is readily estimated, as every 22 gr. of dry carbonic acid gas evolved represents exactly 31 gr. of pure SODA, 47 gr. of pure POTASSA, &c. &c.; these numbers being the equivalents of the respective substances from which the per-centage strength may be found by the rule of proportion, as before explained.
Thus, in the case of a 100-gr. sample of carbonate of soda which has lost 151⁄4 gr. of carbonic acid, by the assay, this would be—
22 : 31 :: 151⁄4 : 21·48%
or nearly 211⁄2 per cent. of pure soda. If 53, the equiv. of anhydrous carbonate of soda, be taken, instead of 31 (the eq. of pure soda), the answer would have been, in the terms of that substance, 36·748%, or nearly 363⁄4 per cent. When an aliquot part of 100 gr. has been taken for the assay, either the result, or the third term of the proportion, must, of course, he multiplied by the denominator and divided by the numerator of the fraction representing such aliquot part.
By multiplying the weight of carbonic anhydride lost, by the numbers opposite the names of the respective alkalies and their carbonates in the second column of the following Table the equivalent per-centage value of the carbonates examined may be obtained in terms corresponding to the various denominations named therein, when 100 gr., or any aliquot part of 100 gr., have been tested; the result, in the latter case, being, of course, multiplied as before.
By taking certain standard weights for the assay, the quantity of carbonic acid evolved may be made to furnish the per-centage strength or value of the specimen in the terms of either the pure or carbonated alkalies, whether in their anhydrous or hydrated state. The numbers in the second column of the following Table represent the quantity in grains and decimal parts of each of the substances named in the first column, equivalent to one grain of carbonic anhydride. These numbers, as already mentioned, may be employed as factors for converting any numbers representing grains of that acid into the equivalents of these substances, true to 4 places of decimals; and further, they furnish us with the data for determining the exact number of grains which must be tested, so that the loss of weight in carbonic anhydride shall at once give us the per-centage richness of the sample in the terms of the denomination for which it is taken. The numbers in the third column of the Table, formed by simply moving the decimal point of the numbers in the second column one figure further to the right, indicate the weights to be taken for the assay, so that the loss of weight, reckoned in tenths of a grain, exactly represents the per-centage
strength in the terms sought. The weights corresponding to the numbers in the fifth column give the same results, provided the loss of weight is reckoned in quarter-grains; those in the sixth column effect the same when the loss of weight is reckoned in half-grains; whilst those in the last column require that the gas eliminated should be counted in grains, and are simply the numbers in the second column of the Table multiplied by 100, or reproduced by moving the decimal point two figures to the right.