rather small. It can, indeed, be increased by the addition of some sulphuric acid, but even then there is much room for improvement.

For the purification of crude anthraquinon it has been recommended to boil the crude article with dilute soda-lye and zinc-dust, to filter hot, and to precipitate anthraquinon from the filtrate by blowing in air. This process has the defect that during the filtration a part of the hydroanthraquinon becomes oxidised, and consequently remains in the residues, and that phenanthraquinon also is reduced by zincpowder and soda-lye. Anthraquinon purified in this manner requires to be treated with sulphuric acid to make it fit for the subsequent operations.

Anthraquinon sublimes in beautiful golden-yellow needles, but on the large scale it is obtained in fine dark gold-coloured columns, several inches in length. The colour varies greatly, depending probably on the size of the crystals. Under certain circumstances, particularly in a state of minute division, it is almost colourless; e.g. if precipitated by water from a solution in sulphuric acid. Even then, however, it is not perfectly colourless, and the golden hue must indubitably rank among its characteristic properties. In ether and benzol it is sparingly soluble, though the latter solvent takes it up more freely at the boiling-point than when cold. Its melting-point is 273° C. Its vapour density was determined experimentally by Graebe as 7.33, whilst, according to the formula C, H,O,, theory would require 7.20. Schützenberger has described an isomeric anthraquinon, which he obtained on repeating the experiments of Graebe and Liebermann. It crystallises in fine red needles, resembling alizarin, from which it is distinguished by its insolubility in caustic alkalis and ammonia. This isomer is transformed into ordinary yellow anthraquinon if its vapour is heated to 300°. It is obtained by the action of nitric or concentrated sulphuric acid upon the products of chloride and anthracen.

8

Auerbach also obtained a red anthraquinon from alizarin paste. If this is evaporated to dryness and sublimed, the sublimate is not perfectly soluble in hydrate of soda; but orange red needles are left behind, completely similar to alizarin in their external properties, but distinguished by their insolubility in caustic alkalis. With zinc-powder and hydrate of soda they give the characteristic red colour of anthraquinon. This red anthraquinon seems to be a compound of anthraquinon and alizarin. If it is treated with benzol, this liquid, on being shaken up with hydrate of potassa, turns violet, whilst the anthraquinon, if treated at once with potassa, shows no colour.

Anthraquinon is distinguished by its stability. It resists the action of oxidising agents with remarkable energy. On fusion with hydrate of potassa it is converted into anthraquinonhydron and anthrahydroquinon.

Bromine has no action in the cold upon anthraquinon, whether dry or dissolved in bisulphide of carbon. On prolonged heating to 160° in sealed tubes bibromanthraquinon is formed.

It is less affected by reducing agents than are the quinons of benzol and naphthalin. In behaviour and in formation the quinon of naphthalin takes an intermediate position between those of benzol and of anthracen. Quinon and its chloro-substitution products are readily reduced by sulphurous acid; bichlornaphthaquinon is not altered by this reagent, but is converted into the corresponding hydroquinon on heating with hydriodic acid. Anthraquinon cannot be converted into bioxyanthracen even by hydriodic acid.

If anthraquinon is heated for some hours to 150° in a sealed tube along with hydriodic acid (boiling at 127°) and some yellow phosphorus, the yellow needles disappear, and colourless leaflets appear in their stead. These, however, consist not of bioxyanthracen, but of anthracen mixed with a little anthracenbihydride. The reduction of anthraquinon

to anthracen is easily effected with zinc-powder. The transformation succeeds perfectly if the anthraquinon is mixed with ten parts of zinc-powder, placed in a combustion-tube closed at one end, and a layer of zinc-powder filled in at the open end. Heat is then gradually applied to the tube from the front to the back, as in an organic analysis, till a faint redness is reached. The anthracen is deposited in the cold front-end of the tube in pure leaflets. As zinc-powder, when heated alone, yields large quantities of hydrogen in consequence of containing hydrous oxide of zinc, the reduction of anthraquinon to anthracen is intelligible :

C11H ̧(O2)+H2+2Zn=C12H10+2ZnO.

Though Graebe and Liebermann stated that anthraquinon resists the action of oxidising agents with extraordinary pertinacity, Wartha succeeded in oxidising it, notwithstanding its remarkable permanence. If an absolute alcoholic solution of pure, almost colourless, anthraquinon (obtained by means of chromic acid, and repeatedly sublimed) is heated to a boil in a test-tube with solid caustic potassa, it is soon found that the liquid turns yellow, and forms two strata-a lower, consisting of melted potassa, and an upper, consisting of an alcoholic solution of anthraquinon. On continued heating the solution becomes darker, and finally, if but little alcohol is present, turns brownish black. The two liquids now mix together with a strong evolution of gas; the mass becomes a fine green, then dark blue, and on continued heating the characteristic violet colour of alizarin-potassa appears. The mixture is now allowed to cool, and dissolved in water; the violet-purple solution is precipitated with sulphuric acid and extracted with ether. From the ethereal solution alizarin, with all its characteristic attributes, can be separated by means of aqueous potassa. In this operation, however, by far the larger part of the anthraquinon is not attacked, and

ACTION OF STANNOUS CHLORIDE UPON ANTHRAQUINON. 73

can be transformed into colouring-matter by repeated fusion with fresh doses of potassa and alcohol.

If stannous chloride is added to the alcoholic solution of anthraquinon, and if it is heated with solid potassa in the same manner to fusion, a peculiar phenomenon is observed. If the chloride of tin is in excess, the lower fused stratum of potassa turns green much sooner than when melted without chloride of tin; the upper alcoholic layer shortly becomes a fiery blood red, and if allowed to cool and to stand uncovered, or if a current of air is passed through, it becomes covered with a brown crust. If this is removed by shaking the liquid, it forms again as long as the fluid displays the intense red colour. The black-brown precipitate deposited, when filtered and washed, can be partially converted into alizarin by solid potassa alone. If the above-mentioned blood-red potassic solution is not allowed to cool, but heated to fusion, the whole mass becomes green, then blue, and finally violet. If an excess of chloride of tin has been added, less alizarin is formed, or sometimes none at all. A small addition of stannous chloride, however, augments the yield of colouring matter. The above-mentioned operations can only be successfully conducted with small quantities. If large amounts are employed, nothing but brown or reddish-brown humus-like bodies are often produced, with little or no colouring-matter. The direct oxidation of anthraquinon can be rendered much more productive by mixing it thoroughly with two parts of sodium-ethylat, and adding it to melting potassa. After the addition of some portions of this mixture there is much frothing, the mass becoming black-brown, and subsequently quite black. The addition is gradually continued, with constant stirring, till the product becomes quite thick, and appears of a blackish-violet in thin layers on the edge of the porcelain capsule. After the mass has been kept at the melting-point for about fifteen minutes, it is allowed to cool; the violet solution is precipitated with sulphuric acid, and

thus the impure colouring-matter is obtained in the form of brown flakes, which are contaminated with undecomposed anthraquinon, as well as with humus-like substances formed during oxidation. It is shaken up with ether, which merely dissolves alizarin and some anthraquinon; the pure colouring-matter is withdrawn from the ethereal solution by means of soda-lye, and the same operations are repeated with the same portion of ether, as long as it takes any colour. The coloured solutions are then mixed together, filtered, washed, dried, and sublimed. The small amount of anthraquinon still adhering to the alizarin may be safely and completely removed by cautiously heating in the sand-bath to 180°–200°. It is only at higher temperatures, approaching 300°, that alizarin sublimes in characteristic orange-red needles.

Graebe and Liebermann have discovered another method of decomposing anthraquinon. If it is heated in a silver capsule to 250° with potassa, the mass soon becomes blue, as if containing alizarin. If, however, water be added, the solution is decolourised, and little is obtained save anthraquinon, which separates out in flakes. If the action is long continued, the anthraquinon is more strongly attacked; on diluting with water, undecomposed anthraquinon is still precipitated, but on adding an acid to the filtrate, there appears a copious colourless deposit, which distils over with the vapour of water, and leaves merely a trace of a brownish colouring-matter (alizarin ?). The acid which separates out in a quantity very large in proportion to the anthraquinon is the benzoic, the formation of which, as the only product of the splitting up of anthraquinon, is the more interesting, as rendering it highly probable that the molecule of anthraquinon is resolved into the same equal halves from which anthracen is formed on its synthesis from benzylchloride. If the temperature of the fused mass of potassa and anthraquinon is raised, a point of time is observed when it is covered with a greenish iridescent film. If the mass is then