in thirty-five teas, consisting of Hysons, Capers, Souchongs, Gunpowders, and others, was 7.67 per cent., the driest teas being the Hysons and Gunpowders, the moistest the Congous : § 201. The Estimation of Theine or Caffeine.-The modern processes for extracting theine fall chiefly under three heads : (1.) Extraction by treating a decoction of the theine-containing substances with lime or burnt magnesia, evaporation to dryness, and subsequent solution of the alkaloid by chloroform, ether, or benzine.— The fundamental idea of this process, perhaps, belongs to Müller; it has also, with various modifications, been recommended by Clous, Commaille, Dragendorff, and many other chemists. Commaille adopts the following method:-5 grms. of the finely powdered and carefully sifted substance are made into a hard paste with 1 grm. of calcined magnesia. This, after standing for twenty-four hours, is dried upon a water-bath and powdered. The resulting green powder is exhausted three successive times in a flask with boiling chloroform, the flask being connected with an inverted Liebig's condenser, so that the action may be continued for a long time. The cool solution is filtered, the chloroform recovered by distillation, and the residue in the flask dried. This residue consists of resinous fatty matters and theine; the former are removed by treating the contents of the flask with hot water and 10 grains of powdered glass, which have been previously washed with dilute hydrochloric acid. The water is boiled and the contents shaken up with the glass; the resinous matters attach themselves to the latter in the form of little globules. The solution is poured on a wet filter, and the residue completely exhausted by repeated boiling with fresh quantities of water. On evaporating the united filtrates in a tared capsule, pure caffeine is left in the form of white crystals. Dragendorff takes 5 grms. of the substance, exhausts it with boiling water, evaporates to dryness, adding 2 grms. of burnt magnesia and 5 of ground glass; the finely powdered residue is soaked in 60 cc. of ether for twenty-four hours, and finally thoroughly exhausted by ether. The latter, when separated and evaporated, leaves the theine in a tolerably pure state. He also states that ether may be replaced by chloroform. Cazeneuve and Caillot recommend a very similar process, but magnesia is replaced by recently slaked lime,* ether by chloroform. Markownikoff uses benzine instead of the solvents mentioned. In all the above processes there is one source of error which does not appear sufficiently guarded against-viz., loss of theine during the evaporation to dryness, since it is absolutely impossible to evaporate a decoction of tea and magnesia to dryness at 100° without loss of the alkaloid-a loss which, so far as the author's experiments go, does not take place until the mixture is quite dry. The following modification may therefore be proposed :--4 to 5 grms. of the tea are boiled in a flask provided with an inverted Liebig's condenser for a couple of hours, the liquid and leaves are transferred to an evaporating dish, some magnesia added, and the whole concentrated to a pasty condition. This paste is treated and thoroughly exhausted by chloroform; the latter is separated and evaporated, and the chloroformic extract redissolved in a little boiling water, the solution filtered, evaporated to dryness at a very gentle heat, and weighed. (2.) Simple Treatment of the powdered Leaves by Solvents.-Legrif and Petit soften the leaves first with boiling water, and then extract the moist mass by the aid of chloroform. Other chemists simply exhaust the powdered substance by chloroform or ether; subsequent purification may, of course, be necessary. (3.) Sublimation.-A method of utilising tea dust by making it a source of theine, was recommended by Heijnsius (Journ. Prac. Chem., xlix. 317). The tea dust was simply treated in a Mohr's benzoic acid subliming apparatus. Stenhouse improved this process by precipitating either a spirituous extract, or a decoction of tea by acetate of lead, evaporating the filtrate to dryness, mixing the residue with sand, and subliming. These processes of sublimation, however, were proposed simply for the extraction, not the estimation, of theine. The writer, in 1877,† proposed the following quantitative method of sublimation :-A convenient quantity of the tea was boiled in the way mentioned, magnesia added, and the whole evaporated to a paste, which was spread on a thin iron plate, and covered with a tared glass funnel. The heat at first was very gentle, but was ultimately raised at the later stages of the process to 200°. The theine sublimes perfectly pure and anhydrous, and forms a coherent white coating on the sides of the funnel; the increase of weight is simply anhydrous theine. To ensure success it is absolutely necessary (1.) That the layer be as thin as possible. * The present writer does not believe that magnesia can be replaced with lime without loss of theine from decomposition. + Op. cit. (2.) That the heat be only gradually increased. (3.) That the mixture be occasionally cooled, and then thoroughly stirred. (4.) That the sublimation be prolonged for a sufficient time. The sublimation is finished when a funnel, inverted over the substance, heated to about 150°, and left for half an hour, shows no crystals. An improvement on this process is to place the paste on a ground glass plate, to which a flanged funnel has been ground so as to fit air-tight. The apparatus is then connected with a Lane-Fox mercury-pump, and an absolute vacuum produced. By the aid of a shallow sand-bath, the theine may be sublimed at a very gentle heat. § 202. Determination of Total Nitrogen.-Peligot, and Wanklyn as well, has laid particular stress on the large amount of nitrogen contained in tea leaves. This nitrogen is, of course, largely dependent on the theine, and it is questionable whether, with the improved methods for the extraction of the latter, it is worth while to make a combustion, more especially as the exhausted leaves are highly nitrogenous, from the presence of an albuminous body. The process is conducted in the usual way in a combustion tube, and best with copper oxide. following are a few determinations of total nitrogen : The Assam tea, from Dr. M'Namara's garden, Mr. Wauklyn has applied his ammonia process to the examination of tea. The soluble matter from 100 mgrms. of tea is heated with a 10 per cent. solution of potash in a flask fitted to a proper condenser, until all the ammonia is distilled over. It may be necessary to add water once or twice, and redistil; then 50 cc. of a strongly alkaline solution of permanganate of potash are added and distilled; the ammonia in the distillates is estimated by "Nesslerising." Mr. Wanklyn gives the following figures as yielded by a genuine tea Free Ammonia, Mgrms. 0.28 0.43 100 mgrms. of genuine tea, sent to the writer by Dr. Shortt, of Madras, yielded total ammonia 81; but this is a method which has not been accepted by chemists, although it has some value. § 203. Determination of Tannin.-The methods proposed for the determination of tannin are very numerous. Four only, however, require any notice here-viz., the gelatine process, the copper process, Mr. Allen's acetate of lead process, and Löwenthal's process. (1.) By Gelatine.-The best process by gelatine is decidedly that which dispenses with the drying and weighing of the precipitate. A solution of gelatine is carefully made by first soaking the gelatine in cold water for twelve hours, then raising the heat to 100°, by placing the bottle on the water-bath (the strength should be about three per cent.), and finally about 8 per cent. of alum should be added. A portion of the solution thus prepared is put into an alkalimeter flask (e.g., Schuster's), and carefully weighed. A solution containing a known quantity of tannin is now titrated with the gelatine until a precipitate no longer occurs; the flask is reweighed, and the loss shows approximately the strength of the solution. One or two more exact determinations will be required to get the correct value. It is necessary to allow the precipitate now and then to settle, and a few drops of the supernatant fluid should be placed on a watchglass, to which a drop of gelatine may be added, and thus the point of saturation ascertained. The tannin in a decoction of tea is, of course, estimated on precisely similar principles. (2.) Copper Process.-When a single determination of tannin is required it is best to precipitate by copper-acetate. 2 grms. of tea are boiled for an hour in 100 cc. of water, the solution filtered, the filtrate boiled, and while boiling 20 to 30 cc. of solution of copper acetate [1:20] added. The precipitate is collected, dried, burnt to an ash, oxidised with nitric acid, and again ignited and weighed. 1 grm. of CuO = 1·3061 of tannin, if Eeler's* figures be accepted; if Woolf's,† then 1 grm. CuO = 1·304 tannin. (3.) Mr. Allen's Lead Process.-A filtered solution of lead acetate 5 per cent., a solution of 5 mgrms. of pot. ferridcyanide, 5 cc. of strong ammonia water, and 5 cc. of pure water, and lastly, solution of pure tannin (1 per cent.) are required. The process essentially depends upon the precipitation of tannin by lead acetate, and using ammoniacal pot. ferridcyanide as an indicator. The latter agent strikes a pink colour with tannin. The solution is standardised by taking a known volume of the lead solution, *Dingler's Poly. Journ., 229, 81. + Zeitschrift f. An. Chem., 1, 104. and dropping in the tannin liquid until a small portion filtered gives a pink colour with the indicator. Tea is tested in a precisely similar manner. Mr. Allen's method is tolerably speedy and accurate; the writer has, however, found the final reaction somewhat difficult to observe. (4.) Löwenthal's Process.-Up to the present time this method (originally worked out for barks) is the best we possess; it depends on the oxidation by permanganate, and indigo is used as an indicator. It not alone gives us the tannin, but the amount of other astringent matters as well. The following solutions are required : (1.) A solution of potass. permanganate, 1·333 grms. per litre. (2.) Precipitated indigo, 5 grms. per litre. (3.) Dilute sulphuric acid (1:3). (4.) A solution of gelatine, 25 grms. to litre, saturated with table-salt.* (5.) A saturated solution of pure salt, containing 25 cc. of sulphuric, or 50 cc. of hydrochloric acid per litre. The analysis as applied to the determination of tannin in barks is performed thus:-10 grms., say, of sumach are taken and exhausted by boiling with water, and the solution made up to 1 litre; of this infusion, 10 cc. are mixed with 75 cc. of water, 25 cc. of the indigo solution added, and 10 cc. of the dilute sulphuric acid. The permanganate solution is run drop by drop from the burette with constant stirring, till the blue colour changes to yellow, when the amount of permanganate used is noted (x). The same process is repeated with indigo and sulphuric acid, and the amount read off (y); subtracting y from x= total astringent matters. The permanganate oxidises both tannin and indigo; but the tannin being the easier to oxidise, is consumed first. In order to obtain accurate results, the proportion of indigo should be such as to require about twice the quantity of permanganate which would be consumed by the tannin alone. Thus, if indigo alone requires 10 cc. of permanganate to decolorise it, the indigo and tannin together must not take more than about 15 cc.; if it does so, the tannin must be diluted accordingly. The total astringent matters being known, the next step is to throw the tannin out, and estimate the gallic acid and impurities. 100 cc. of the infusion are mixed with 50 cc. of the salted gelatine infusion; after stirring, 100 cc. of the salt acid solution are added, and the mixture allowed to stand Löwenthal prepares the solution by steeping 25 grms. of the finest Cologne glue in cold water over night; it is then melted on the water-bath, saturated with NaCl, and made up to 1 litre with saturated NaCl solution, filtered, and kept well corked. |