Before use the berries are roasted to a chocolate brown, and then ground in a mill.

Microscopical Structure.-The main portion of the berry is composed of strong, angular, thick-walled cells, as represented in fig. 39, a. The figure is an accurate representation of what was seen in a fine section of a sample of unroasted coffee, but this particular sample shows the cell divisions much more distinctly than usual. When the coffee has been roasted, the usual appearance of these cells is that of very dark, small, irregularly-triangular masses, representing the contracted cell-contents set in a transparent matrix, in which it is difficult to trace the cell-walls. The cellular elements of the seed, more especially in the outer layers, are impregnated with drops of oil. A thin, tough, Japanese-paper-like membrane (fig. 39, b) may be detached from the berry, and is found to be composed of a series of spindle-shaped fibres attached to a tissue which shows signs of fibrillation, and is probably composed of a number of very fine fibres, adhering by their edges, so as to form a continuous sheet.

The microscopical structure just described separates and distinguishes coffee at once from all other known berries or seeds, while the tissues of roots, such as chicory (consisting of loose, thin-walled, vegetable cells, with a greater or less admixture of large spiral vessels), are entirely different, and may be said, indeed, to be built on a different plan.

§ 211. Chemical Changes during Roasting.-The effect of roasting is to drive off a large quantity of water, to volatilise a small quantity of theine, to change a portion of the sugar into caramel, to rupture the cell-layers containing fat and albumen, and to swell the berry by the extrication of gases, consisting mainly of carbon dioxide. There is also developed a fragrant aromatic substance, a single drop of which is sufficient to scent a large room with the peculiar coffee odour; the best temperature for the production of this aroma is 210°. That during roasting there is an actual loss of theine, is easily proved by holding a glass plate over the heated berries; in a very little time crystals of the alkaloid condense. Tenneck found in unroasted coffee ·75 per cent., and in the same roasted, 42 per cent. theine. It would

appear that roasted coffee gives up more to water than does raw coffee; for Cadet found that beans roasted to a light brown yielded to water 12.3 per cent.; to a nut-brown, 15.5 per cent. ; and to a dark-brown, 21.7 per cent. Vogel also gives the soluble matters in raw coffee as 28 per cent.; and in roasted, 39 per cent. The amount of sugar changed is always considerable; thus, Graham and Stenhouse found the following differences in the percentage of sugar between the raw and roasted coffees:—

König has studied the changes taking place in roasting, and his views and experiments are as follows* :

-300 grms. of coffee-berries containing 11-29 per cent. water gave, on roasting to a light brown colour, 2467 grms. of roasted coffee, containing 3.19 per cent. water. We have therefore

Hence, in roasting, 866 per cent. of water, and 9.11 per cent. of organic substances have been lost. This is divided among the constituents of coffee as follows. (See Table XXV.)

§ 212. Constituents of Coffee.-The main properties of coffee are apparently due to four distinctive substances:-(1.) Unessential oil, not yet completely studied; (2.) caffeo-tannic acid; (3.) theine or caffeine; (4.) a concrete oil or fatty substance.

Theine, or Caffeine, is described at p. 316.

Caffeo-tannic Acid, CHO,, was first observed by Pfaff, in the seeds and leaves of the coffee plant; it also occurs in the root of the Chiococcu racemosa, and in the leaves of the Ilex paraguayensis, S. Hil. It may be separated from coffee by fractional precipitation of the infusion with acetate of lead. The precipitate, at first falling, consists of citrate with caffeo-tannate of lead; but later on, the latter occurs alone, and can be washed with water and decomposed by SH, in the usual way. Caffeo-tannic acid thus obtained is a brittle, yellowish mass, easily powdered, and of feeble acid reaction. It is supposed to exist in the plant in combination with potash and theine. It is scarcely soluble in ether, but dissolves easily in water or in alcohol. The solution gives a

*Nahrungs-und Genussmittel, Bd. ii., 479.

TABLE XXV.

+(0'078)?+(0·74)

- 10·9% +

+? +? -3·1% -2·1% - 66·9% +3·9% -280°-57-76·7%

The percentage composition of the dry substance was as follows:

dark-green colour with chloride of iron, or if dissolved in aqueous solution of ammonia, the alkaline earths, or the alkalies, a reddish-yellow or yellow colour. If caffeo-tannic acid be decomposed with 3 parts of solid potash, the end product is protocatechuic acid. If it be boiled with 5 parts of potash-lye (1.25 specific gravity) for three-quarters of an hour, and then neutralised by sulphuric acid, caffeic acid (C,H,O,) crystallises out, and can be obtained in straw-yellow prisms and plates by recrystallisation from hot water. It colours chloride of iron grass-green, is oxidised by nitric acid into oxalic acid, and is decomposed by potash into protocatechuic and acetic acids.

Coffee Fat.-The coffee fat can be obtained from an alcoholic extract of coffee; part separates on cooling the fluid to 0°, the rest on dilution with water. It is white, without odour, of a buttery consistence, melting at 37°5, and becomes rancid on exposure to the air. According to Rochleder (Wien Akad. Ber., xxiv. 40), it contains the glyceride of palmitic acid and of an acid of the composition C12H2402

C. O. Cech exhausted 50 lbs. with aloohol and ether, and obtained 1,200 grms. of a thick green fluid oil, and after a time fine crystals of theine separated. After six months the oil, although in closed flasks, began to be turbid, and gradually little groups of crystals separated and sank to the bottom. After the lapse of three years, the flask was found to be about two-thirds filled with crystals of the more solid fats, but the upper layer was beautifully green.

There are also citrates, and probably other organic acids and nitrates in coffee. J. Buig found in raw dry coffee 054 per cent. of nitrate of potash, in roast 041 per cent.

The following table gives the general composition of various commercial varieties of coffee :

Finest Jamaica Plan-) 25.3 1.43 14.76 22.7 33.8 3.8

tation,.

1.87 0.31

PhosAsh. Potash. phoric

Acid.

Finest Green Mocha,

Some analyses of Dragendorff may be also quoted:-*

§ 213. Analysis of Coffee.-The hygroscopic moisture, theine,

* There is also a very elaborate analysis of coffee made by Payen, at a date when food analysis was not so well understood as now. It is probably a fair approximation as regards the more important constituents, but the percentage of ash can hardly be correct.