parts of water, at a low temperature. The melt soon takes a yellowish red colour, and on the addition of an acid gives a yellow or orange precipitate, consisting of oxyanthraquinon mixed with a little alizarin. If less alkali and more water be taken, a yellowish-red solution is still obtained, which on the addition of acid deposits a grass-green precipitate which dyes silk and wool direct of a green shade, and consists probably of hydroxyanthraquinon.

In order to separate the oxyanthraquinon from the accompanying alizarin, the mixture is best boiled with carbonate of lime, baryta, or lead, when soluble salts of oxyanthraquinon and insoluble alizarin salts are formed. The mixture is filtered, and oxyanthraquinon is separated from the filtrate by the aid of acids. To obtain it pure these operations are repeated several times, and the substance is then sublimed or recrystallised from alcohol. On sublimation it is obtained in beautiful lemon or orange-coloured crystals. From alcohol or ether it crystallises in fine yellow needles.

This method of separation is not, however, quite accurate, since, on the one hand, oxyanthraquinon-calcium is not very soluble in water, and alizarin-calcium is not quite insoluble. Hence on extracting the lime-salts in boiling water a certain quantity of alizarin-calcium passes into solution. Willgerodt, therefore, makes use of another method which, according to his statement, can be used as an approximate quantitative separation of the two bodies. He dissolves the mixture of alizarin and oxyanthraquinon, after the acids, salts, and anthraquinon have been removed by well-known methods, in hydrate of potassa dissolved in water and exactly calculated for the quantity employed, so that a perfectly neutral alizarate may be formed. He evaporates to dryness in the water-bath, and repeatedly extracts the finely pulverised residue with alcohol, until the solution, at first bloodred, takes a brownish-violet colour. The extracts are poured

together and evaporated to dryness, the residue again dissolved in water, when, on the addition of an acid, oxyanthraquinon separates as a straw-coloured gelatinous precipitate. The portion insoluble in alcohol is pure alizarin. The oxyanthraquinon thus obtained may be recrystallised from glacial acetic acid. It is insoluble in boiling alum-water, and sparingly soluble in boiling water.

This method presupposes an analysis of the mixture of alizarin and oxyanthraquinon, as it would be otherwise impossible to calculate the exact weight of hydrate of potassa to be employed.

Oxyanthraquinon is scarcely soluble in cold, and sparingly in boiling water; somewhat readily in alcohol and ether. With bases and carbonates it behaves like alizarin; but it has only one atom of hydrogen capable of being replaced by metals. Its salts are soluble in water with a yellowish-red colour. With concentrated sulphuric acid there is formed at low temperatures a brownish-red solution, from which oxyanthraquinon is reprecipitated unchanged on the addition of water. If the solution in sulphuric acid is heated more strongly until a precipitate no longer appears on the addition of water, a sulpho-compound is formed, not, however, identical with the ordinary sulphoxyanthraquinonic acid, but soluble in alkalis with a red-yellow colour. From boiling glacial acetic acid, oxyanthraquinon crystallises in long yellow needles. It does not dye mordanted cotton, and its melting-point is 323°.

The above-mentioned transition into alizarin is a characteristic of oxyanthraquinon. If heated with zinc powder it yields anthracen.

7

The baryta salt, [C,,H,(02)0], Ba, separates from the hot saturated aqueous solution in yellow microscopic needles. It is not very freely soluble in water, but more readily when hot than when cold; it is insoluble in alcohol. The barytic salt of oxyanthraquinon cannot be obtained by concentration.

ERYTHROXANTHRAQUINON AND OXYANTHRAQUINON. 107

If we attempt to remove the excess of baryta from its solution in baryta water by means of carbonic acid, the reddish-yellow solution is completely decolourised, the salt being decomposed into carbonate of baryta and free oxyanthraquinon. The yellow precipitate which contains both substances yields, on boiling with water, a red solution of oxyanthraquinon barium, since the oxyanthraquinon with the aid of heat expels the carbonic acid of the BaCO3. In the same manner it decomposes marble.

When a boiling concentrated solution of oxyanthraquinon in hydrate of baryta is allowed to cool, a light-red crystalline barytic salt is deposited. Oxyanthraquinon forms with cold concentrated baryta water a perfectly clear solution, but after a few minutes it separates almost completely as a red barytic salt. This compound is probably a basic salt of monoxyanthraquinon. If we attempt to wash it, it chiefly dissolves, and the washing water has a strongly alkaline reaction.

If mixed with alizarin, it communicates a red colour to the alkaline solution of the latter, whence it was formerly believed that alizarin contained purpurin, which has not yet been found in natural alizarin.

is obtained by heating oxyanthraquinon to 160° with anhydrous acetic acid. On recrystallisation from alcohol it forms small felted yellow needles. With alcoholic potassa it remains, in the cold, colourless. If heated together, the abscission of the acetyl group can be very accurately traced, as the solution very soon assumes, with increasing intensity, the reddish-yellow colour of oxyanthraquinon potassium. Ultimately pure oxyanthraquinon can be precipitated by acids.

Erythroxyanthraquinon, CH,O3.

As was mentioned in speaking of oxyanthraquinon, there

are formed by the reaction of phthalic acid and phenol two isomeric bodies--oxyanthraquinon and erythroxyanthraquinon. The latter is almost insoluble in dilute ammonia, and remains therefore undissolved on the treatment of the crude product of the action of phenol upon phthalic acid. To obtain it in a state of purity, the portion insoluble in ammonia is dissolved in hot alcohol and mixed with baryta water. The deep-red precipitate formed is washed and decomposed with hydrochloric acid, and the yellow precipitate is recrystallised from boiling alcohol. On cooling, the entire liquid is filled with yellowish-red dendritically entangled needles, which, when dry, assume a beautiful reddishgolden lustre.

The properties of the two oxyanthraquinons differ decidedly from each other in many respects, as will appear from the following tabular comparison :—

Erythroxanthraquinon.

Crystallises from alcohol in agglomerated orange needles, more readily soluble in hot than in cold alcohol.

Melts at 1730-180° C.; begins to sublime at 150°; and condenses in long orange needles of the shade of sublimed alizarin. If suddenly heated in a test-tube the vapour is deposited in oily drops.

Easily soluble in dilute potash-lye with a reddish-yellow colour.

Almost insoluble in dilute ammonia; very sparingly soluble in concentrated ammonia, with a reddish colour.

Yields with baryta and lime-water a deep-red lake; almost insoluble in water, decomposable by carbonic acid, and does not decompose carbonate of baryta on boiling in water.

Forms no coloured compound with alumina.

Soluble in alcohol and ether with a yellow, and in concentrated sulphuric acid with an orange colour.

Oxyanthraquinon.

Crystallises from alcohol in single needles or leaflets of a pure sulphur yellow; not much more readily soluble in hot than in cold alcohol.

Melts at 268°-271° (according to Willgerodt, 323°), and sublimes at a higher heat in yellow leaflets, as also, if suddenly heated, in a test-tube.

The same.

Easily soluble in ammonia with a reddish-yellow colour.

Readily soluble with a reddishyellow colour in baryta and lime-water; dissolves carbonate of baryta if boiled in water.

The same.

The same.

The absorption spectrum of the alkaline solution offers nothing remarkable; but there is a band in the sulphuric solution.

No band in the absorption spectrum of either the alkaline or the sulphuric solution.

Professor Kundt has compared the optical properties of the two compounds, and gives the following report on the behaviour of their solutions in concentrated sulphuric acid :

Steinheil's Spectrum Apparatus; soda line at 172°.

Erythroxanthraquinon.
(Medium Concentration.)

The blue part of the spectrum quite absorbed. Boundary of the absorption somewhat sharp at 192°; then a dim but distinct absorption band in the green at 185°.

(Very weak Concentration.) Blue not entirely absorbed; the maximum absorption about 120°.

Oxyanthraquinon.
(Medium Concentration.)

Blue part of the spectrum quite absorbed. Boundary of the absorption at 180° not so sharply defined. No band can be seen.

(Very weak Concentration.)

The extreme blue part of the spectrum not absorbed; maximum absorption about 120°.

Erythroxyanthraquinon is therefore readily distinguished from oxyanthraquinon, which it otherwise much resembles, by the presence of an absorption band in its solution in sulphuric acid.

Erythroxanthraquinon, by the formation of baryta and lime salts, approximates much more closely to alizarin than does oxyanthraquinon. The position of the hydroxyls, however, in both corresponds to the position of each hydroxyl in alizarin. On melting with alkali, erythroxyanthraquinon yields alizarin, but it fuses less easily. The resulting alizarins are identical in their optical behaviour.

Sulphoxyanthraquinonic Acid, C14H6(O2)(OH)(SO,H). The formation of this acid was mentioned when treating of bisulphanthraquinonic acid. To obtain it we proceed as follows:-Bisulphanthraquinonic acid is heated till the blue