| Tannin (as Oxalic Acid). | Other Bodies Oxidised (as Oxalic Acid). | |
| Spent Liquor | 0·12 | 11·0 |
| Valonia (good Smyrna). Sample 1 | 29·1 | 2·3 |
| " " Sample 2 | 30·7 | 2·1 |
| " " Sample 3 | 30·5 | 1·9 |
| Hungarian Larch Extract. Sample 1 | 14·78 | 1·95 |
| " " Sample 2 | 18·08 | 2·33 |
| Chestnut-wood Extract, 25° B. | 25·53 | 3·68 |
| Pegu Cutch | 63·59 | 2·45 |
It is proved by experiment that kaolin removes nothing which is oxidised by permanganate, but simply facilitates the precipitation and filtration; and it is often found useful to clarify the original infusions and liquors before the first titration. On the other hand, there is no doubt that the salt and acid of Löwenthal's method precipitate of themselves a large proportion of certain tannins. In the case of cutch this amounted, in the analysis given, to 67 per cent. of the whole. There is, however, good reason to believe that this would also have been absorbed, or at least removed from solution by hide in the process of tanning. This is shown by the analysis of the spent liquor above given, which originally contained the tannins of oak bark, valonia, myrabolans, gambier, hemlock, and oak wood extracts, &c., to the extent of 10 to 15 per cent., but which was reduced by contact with hide to 0·12 per cent. That a portion had not been absorbed but decomposed is proved by the large accumulation of oxidisable impurities (equal to 11 per cent. of oxalic acid); at the same time this example shows that the method is capable of estimating a very small portion of tannin in presence of much gallic acid and other analogous substances. It is worth remark that such spent liquors become very pale in colour, and also that the filtrates, freed from tannin by precipitation, are nearly colourless, thus proving that the colouring matters present in tanning materials are of the nature of tannins, at least as regards their precipitability by hide and gelatin.
Simand (Dingl. Polyt. Jour., ccxlvi. 133) has recommended instead of precipitation with gelatin, the use of the gelatinous tissue of bones to remove the tannin. For this purpose porous bones, such as horn piths, are coarsely powdered, and after treatment with dilute soda solution to remove the fat, are steeped in weak hydrochloric acid till all the calcareous matter is dissolved. They are then thoroughly washed, ground wet through a steel mill, washed again and dried at a low temperature; the tannin is removed more quickly than by raw hide, and the amount of gelatinous matter dissolved by cold water is a very trifling one. This method, or that with purified hide-powder, is to be recommended for scientific research, since no element capable of precipitating substances other than those absorbed by the hide is introduced, while it is not certain in all cases that saturation with salt and acidification may not remove other constituents of the liquor besides tannins. It has, however, for technical purposes the great disadvantage of requiring a much longer time for absorption of the tannin than is the case with gelatin solution, and of the process being much more difficult of execution. If hide-powder be employed, it must be moistened with a small quantity of water before adding to the infusion, and this water must be taken into account in the quantity of the filtrate employed for the titration of the "non-tannin." The digestion with the hide- or bone-powder must be continued till the filtered liquid does not give the faintest clouding with a drop of clear gelatin solution, and it is always very difficult to be sure that the tannin is so completely removed as with gelatin and salt. Hide- or bone-powder may be employed to determine the actual weight of any unknown tannin absorbable by hide, by evaporating equal quantities of the original infusion and of that freed from tannin by digestion with the powder; the difference giving the tannin absorbed. The evaporation must be conducted as far as possible in absence of air, for instance in vacuo, or in a current of carbonic dioxide, and the residues both dried at 212° F. (100° C.) so long as they lose weight. The amount of matter dissolved from an equal quantity of the hide- or bone-powder by water must also be ascertained and taken into the calculation.
Ammoniacal solution of cupric acetate or sulphate has been employed by several chemists to remove tannin from solutions. N. H. Darton of New York, who has a large practice in tannin analysis, employs cuprammonic sulphate in the following manner.
The infusion, for which 20 grm. of hemlock bark or a corresponding quantity of other material must be used, is made by exhausting with 2 or 3 quantities of water successively, first cold, and then with heat (by placing the flask in a pan of boiling water), each portion of water being poured off into a litre flask. The last should be almost colourless. The liquor is thus made up to nearly 1 litre, 25 c.c. of dilute sulphuric acid (about 1 vol. concentrated in 10) is added, and the liquor is filtered through a small filter, which is finally rinsed with a small quantity of water. Liquid ammonia is now added till the liquor slightly smells of it, and, if any precipitate is formed, it is filtered off as before; 25 c.c. of dilute sulphuric acid is again added (which should give the liquid an acid reaction), and it is made up to 1 litre. The titration is done as described under Löwenthal's method, but instead of precipitating with gelatin, 100 c.c. is mixed with 100 c.c. of a solution of copper sulphate to which sufficient ammonia has been added to redissolve the precipitate first formed, and containing 11/4 per cent. of copper sulphate. This is well shaken and filtered, and the "not-tannin" is determined in the filtrate just as with gelatin; a little dilute sulphuric acid being added in the basin to neutralise the ammonia. The writer has examined this method with regard to a few tanning materials. With valonia (and therefore probably with oak bark) the preliminary treatment is unnecessary, and copper precipitation gives results practically identical with the improved gelatin, while it is less troublesome. On the other hand, a sample of Miller's Hungarian Larch Extract which gave tannin equal to 18·08 per cent. (by the gelatin method) gave no precipitate with cuprammonic sulphate, and hence a result in tannin of nil by Darton's method. It is worth remark that by the copper method it is therefore possible to estimate the valonia tannin alone in a mixture of larch and valonia tannin. Probably this mode of analysis may also be utilised to separate other tannins. With chestnut extract the results seem satisfactory, as regards the precipitation of the tannin by copper, the figures agreeing very closely with those by gelatin, but the preliminary treatment with sulphuric acid and ammonia precipitates about 75 per cent. of what is usually reckoned as tannin, leaving 7·53 per cent. of tannin only instead of 25·53 per cent. as reckoned by the gelatin method; which, judging by practical results in tanning, can hardly be accepted as correct. The results of the gelatin method are found to agree fairly with those of direct absorption by hide-powder, which is strong confirmation that what is estimated as tannin is what is absorbed by the hide. It is well known that sulphuric acid precipitates many tannins, and in an experiment with cutch it was found by the writer that saturation with salt and the addition of dilute sulphuric acid as for Löwenthal's process, but without the gelatin, precipitated 67 per cent. of the total tannin as usually reckoned.
It is obvious that it is impossible by analysis to compare the relative value of different tannins, such as those of myrobalans and gambier, or hemlock and valonia. All that analysis can reasonably be expected to do is to give the relative values of different samples of the same substance, or at the most, of materials of the same class. All other comparisons are misleading; and would be so, even if the exact percentage of each tannin could be calculated; since the commercial and practical value of different materials does not depend on the quantity of tannin only, but on the character of the leather it produces, hard or soft, dark- or light-coloured and heavy- or light-weighing.
A Commission of German technical chemists, under the presidency of Dr. Councler of Eberswalde, and including Messrs. Eberz, Kathreiner, Schaun, von Schroeder, and Simand, have recently reported on methods of tannin estimation ('Bericht über die Verhandlungen der Commission zur Feststellung einer einheitlichen Methode der Gerbstoffbestimmung,' Cassel, 1885). After reviewing earlier methods, they recommend the following modifications of the Löwenthal method, for general adoption.
Chemicals employed.
(1) Permanganate solution. 10 grm. of the purest potash permanganate are dissolved in 6 litres of distilled water.
(2) Indigo solution. 30 grm. dry sulphindigotate of soda (Carminum cærul. opt., "pure Indigotin I" of Gehe & Co., Dresden), air-dry, are dissolved in 3 litres of dilute sulphuric acid (1 vol. H2SO4 to 3 vols, water), 3 litres of distilled water are added, the whole is shaken till dissolved, and filtered. In each titration, 20 c.c. are used in 3/4 litre of water, and reduce about 10·7 c.c. of permanganate.