on all sides by a network of round or oval cells filled with chlorophyll grains (Fig. 4, A). Scattered amongst these cells are peculiar branched bodies which have been sometimes called "branched and spinous hairs" (Fig. 4, c), but which are, in reality, branched, thick-walled cells. There are also

present in the leaf numerous sphæraphides and other plant crystals.

Chief Characteristics.-These are the looped or meshed venation of the leaf, the dentate margin with the short spines, the slightly emarginate apex, the numerous stomates, the unicellular hairs, and the peculiar branched cells referred to.

ANALYSIS.

Although, in the analysis of tea, the more characteristic principles have chiefly engaged the attention of chemists, a more or less complete analysis of the leaf has been attempted. Various methods appear to have been adopted to arrive at an accurate estimation of the several constituents; but, from the very different results obtained, some of them must obviously be incorrect.

Instead of making a water extract only, we have first exhausted the tea as completely as possible with alcohol of 70 per cent., and then with water, keeping the extracts apart and examining them separately. The result of following this process is that a larger total extractive is obtained than by water only, nearly all the tannin and soluble nitrogenous matter being found in the alcoholic extract, while at the same time the constituents soluble in water can be more readily dealt with.

Oil. This is determined by distilling about 100 grains of the tea in presence of water. The distillate possesses a strong odour of tea, and contains a little chlorophyll and resinous matter mechanically carried over, which should be removed by filtration. The filtrate is saturated with chloride of calcium and shaken with ether in a separator. The upper layer of liquid is then drawn off into a tared beaker, and the ether evaporated at a temperature of about 80° F. (26.6° C.). The weight of the beaker and oil is then ascertained.

Theine. One of the two following methods is usually adopted for the estimation of theine in tea :

First Method. One hundred grains of dry and powdered tea are boiled for several minutes with an equal weight of calcined magnesia and 8 ounces of strong alcohol, and filtered. The boiling is repeated with a like amount of alcohol, and subsequently three times with distilled water, filtering after

each operation. The alcohol is recovered by distillation from the united alcoholic filtrates, and the residue, after addition of water, is filtered to remove colouring matter thrown out of solution. A further amount of colouring matter is got rid of by evaporating this filtrate to dryness, and exhausting the residue with boiling water. The resulting solution is added to the aqueous extracts obtained directly from the tea, and the whole evaporated to dryness with the addition of a further small quantity of magnesia, and completely exhausted with pure hot benzol. On distillation, or evaporation of the benzol, the theine is left in a crystalline and almost colourless state.

Second Method.-One hundred grains of dried tea are reduced to a powder, and thoroughly exhausted with boiling water. To the extract, subacetate of lead is added as long as a precipitate appears. The precipitate is then removed by filtration, and the lead in the filtrate got rid of by means of sulphuretted hydrogen. The filtrate from the sulphide of lead is gently evaporated to a small bulk, and left for twenty-four hours, when the theine will be found to crystallize out in an impure state. The crystals are removed from the mother liquor, pressed between bibulous paper, dissolved in a little water, and left to re-crystallize. The mother liquor still contains a little theine, which some chemists have endeavoured to estimate by a standard solution of tannic acid. In the course of our experiments we have found the first of these methods to yield the best results.

Tannin. The amount of tannin in tea is sometimes roughly estimated by weighing the precipitate occasioned by a solution of gelatine and alum in a given quantity of tea solution and reckoning 40 per cent. of the dried product as tannin. By a second method 4 grams of gelatine and 2 grams of alum are dissolved in warm water and made up to 1 litre ; then I grain of the finest tannic acid is dissolved in 40 cubic centimetres of water and 10 cubic centimetres of a saturated solution of common salt, added along with a little ground glass. The gelatine solution is introduced

C

into a burette, and run into the solution of tannic acid until no further precipitate forms. The number of cubic centimetres is read off and noted.

Twenty-five grains of the powdered tea are next exhausted with boiling water, the solution filtered and made up to 200 cubic centimetres. 40 cubic centimetres of the solution (= 5 grains of tea) are transferred to a beaker, and solution of salt and some ground glass added as above. The process is now conducted as in the test solution. There appears to be an advantage in making the liquid in the beaker specifically heavier than the gelatine solution, as the latter, when carefully dropped into the beaker, remains for a short time on the surface, and thus allows any formation of precipitate to be more easily discerned.

Suppose that the one grain of tannic acid requires 14 cubic centimetres of gelatine solution, and 40 cubic centimetres of tea solution require 9.8 cubic centimetres of the gelatine, then 14: 1: 9.8 × 20= 14 per cent. of tannic acid in the sample of tea.

The following method has been proposed by Mr. Allen. A solution of acetate of lead in distilled water is prepared of a strength equal to 5 grams in the litre. 1-10th gram of pure tannic acid is dissolved in 100 cubic centimetres of water, and introduced into a burette; 10 cubic centimetres of the lead solution are measured into a beaker, and diluted with 90 cubic centimetres of water. This solution is made hot, and the tannic acid solution carefully dropped into it until a little of the filtered liquid gives a pink colouration with a drop of an ammoniacal solution of ferricyanide of potassium.

The tea solution is prepared by exhausting 2 grams of the powdered tea with boiling water, and making up to 250 cubic centimetres. This solution is tested in the same way, and the calculation of the percentage of tannin made from the two results. It is evident that some precautions are necessary in applying the test, as there is no provision made here for the precipitation of the lead by the alkaline phosphates and other substances pre

:

sent in the solution which are thrown down by a salt of lead.

Another method by Löwenthal, with modifications by Estcourt, has been suggested. For this test various solutions are required: 1. Solution of permanganate of potassium made by dissolving 1.66 grams of the salt in a litre of water.

2. Solution of pure sulphate of indigo: 30 grams dissolved in water and made up to a litre.

3. Solution of pure tannic acid: 2 grams in 1 litre of distilled

water.

4. Dilute sulphuric acid: 200 cubic centimetres of concentrated sulphuric acid diluted to 1 litre, or 1 in 5.

5. Solution of pure gelatine: 25 grams of best glue dissolved in warm water, and made up to 1 litre with a saturated solution of common salt.

In the application of the test, the tannic acid, indigo, and other matters in the tea solution are oxidised by the permanganate, and the end of the process is shown by the change of the blue colour to green and then to a pale yellow.

The first point to be ascertained is the number of cubic centimetres of permanganate solution required to decolourise 20 cubic centimetres of the indigo solution. For this purpose 20 cubic centimetres of the solution are diluted with 800 cubic centimetres of distilled water and 10 cubic centimetres of the sulphuric acid solution added. The permanganate solution is run in from a burette until the blue colour completely disappearsthe end of the process being carefully observed. The same quantity of indigo solution is again taken, with the addition of 10 cubic centimetres of the tannin solution, and the permanganate solution added as before. The difference in the two results is due to the tannic acid. It is desirable that the experiments should Occupy the same time-about 6 minutes-and the solution should be kept vigorously stirred during the addition of the permanganate.