A convenient plan of estimating the tartaric acid volumetrically, is to dissolve the precipitate in hot water and titrate the solutiori volumetrically. This plan gives results closely according with those obtained by direct weighing when pure tartaric acid is employed.
Another series of experiments was conducted in a similar manner, except that the precipitate was washed with proof spirit, which had not been previously saturated with acid potassium tartrate. The following results were obtained, no connection being made for solubility of the precipitates in the mother liquor:— These estimations were very satisfactory when the proportion of tartaric acid did not exceed 20 or 30 per cent., but there Was a uniform tendency towards too high a result. The discrepancies observed seem attributable to (1) the presence of tartaric acid in the sample of citrate employed ; (2) precipitation of an acid citrate together with the potassium hydrogen tartrate.
Apart from direct experiment having proved the absence of tartaric acid iu the original citric acid used, the invalid nature of the first explanation is shown by the fact that the less citric acid used the grazier was the excess of tartaric acid found over that added.
Experiments were then made with the view of ascertaining whether the discrepancy was due to the second cause. A quantity (2 grammes) of the citric acid was treat cd in exactly the usual manner, and gave no evidence of the presence of tartaric acid. On the following morning the liquid was again stirred, the temperature being only about 8° (46° F.), when well-defined streaks were produced in the track of the glass rod, and in a few minutes the liquid became semi-solid from the formation of a crystalline precipitate. Either the sample was largely contaminated with tartaric avid, or the cold had induced the separation of potassium citrate. The latter, as might be ex pectol, proved to be the truth ; warming causing the precipitate to disappear gradually, while, on decantiiig off the alcoholic liquid and adding a moderate quantity of cold water, the precipitate dissolved instantly. The fact deserves especial attention, as ignorance of it might readily lead to a sample of citric acid being condensed as largely adulterated with tartaric acid, when the latter substance was entirely absent.
The above observation naturally gave the clue to the anomalous results already obtained. In the concentrated and highly alcoholic solution used, there was a tendency to the precipitation of potassium citrate along with the tartrate. Thia tendency seemed capable of correction by treating the washed precipitate with a cold saturated aqueous solution of petas:inta hydrogen tartrate, which would readily dissolve any precipitated citrate, without acting on the acid tartrate.
The following figures show the results of this treatment: In this case, the results obtained from the sample containing 20 per cent. of tartaric acid were almost within the limits of error, while the reduction of the weight el the precipitate in No. 2 by an amount equal to nearly 6 per cent. of tartaric acid, conclusively shows that the previous excess was duo to citrate carried down by the tartrate precipitate. If, to the results obtained in No. 2 wo add 1.1 per cent., as correction for solubility of the potassium hydrogen tartrate in the mother Uglier, we obtain per cent. of tartaric aoid found, as against 40 per cent. added.
An attempt was next made to obtain a precipitation in an acouotis solution, using proof spirit merely for washing the product. Two quantities dissolved in ten times their weight of water gave by this method and per cent. of tartaric acid respectively, instead of 40 per cent. and 20 per cent. added.
Next aome experiments were made, in which a cold saturated aqueous solution of potasaium hydrogen tartrate was used as the solvent of the sample, but the results were unsatisfactory.
The results of the whole of the above experiments have led to the use of the following iiroo:.ss, which, while readily detecting 2 or 3 per cent. of tartaric acid, allows of the estimation of larger proportions with very fair accuracy :—Dissolve 2 grammes of the sample to be tested in 45 ce. of proof spirit, filter from any undissolved calcium or potassium tartrate, add 5 cc. of a cold saturated solution of potassium acetate in proof spirit, stir, and allow to stand for 12 hours. Filter off the precipitate produced and wash it with proof spirit. Rinse off the precipitate from the filter with a saturated solution of potassium hydrogen tartrate in cold water, digest in the cold for a few hours with occasional stirring, then filter, wash ouce with proof spirit, rinse off the precipitate into a small porcelain dish with boiling water, evaporate at 100°, and weigh the acid potassium tartrato obtained. The weight, multiplied by (or roughly, gives the quantity of tartaric acid in 2 grin. of the sample examined. Aa a check, the dry precipitate may be ignited, and the solution of the ash titrated with standard acid ; in the same method (substituting standard alkali for acid) may be applied to the purified precipitate on the filter, so as to avoid the trouble of the subsequent evaporation at steam heat.