perhaps belonging to the dicarbonic acid, remains behind. The re-precipitated acid is again treated in the same manner, and then crystallised from alcohol.

Anthracencarbonic acid is insoluble in water, sparingly soluble in benzol, more readily in alcohol, ether, and glacial acetic acid. From alcoholic solutions, saturated when hot, it is deposited on cooling in fine yellow needles; if the concentration is more rapid, it appears almost amorphous. The acid decomposes carbonates, and forms chiefly soluble salts. The barium and calcium salts dry up to amorphous, pale yellow masses, which become anhydrous at 130° C., and have then the composition (C,,H,CO2),Ba, and (C,,H,CO2),Ca. The sparingly-soluble lead salt is thrown down in flakes on mixing the solution of the ammonia-salt with a solution of acetate of lead. The barium-salt is not decomposed by carbonic acid, but the ammonium-salt loses its base on evaporation. The solutions of the acid and its salts fluoresce from greenish to blue, according to their degree of purity. The acid is always precipitated of a yellow colour, even from solutions free from the slightest yellow tint.

If anthracencarbonic acid is heated with soda-lime, or if its lime or baryta-salt is heated alone, anthracen is obtained.

So far this acid and its salts possess an external resemblance to the anthracencarbonic acid described by Graebe and Liebermann, and obtained by another process. A slight distinction appears on crystallising the acid from alcohol, since the earlier-described acid, on the evaporation of the solvent, is deposited in long needles.

Considering our slight acquaintance with the monosubstitution-products of anthracen, it seems important to decide whether the two anthracarbonic acids are identical or isomeric, since in the latter case we may infer the possibility of isomeric monosubstitution-products in general.

The constitution of anthracen at present assumed

would admit of three series of isomeric monosubstitution products (in a, b, and c). The experiments of Liebermann and V. Rath have proved the isomerism of this acid with that of Graebe and Liebermann.

The essential distinctions are as follows:

The acid of Graebe and Liebermann melts at 206° C., but passes partially even at 150° C. into anthracen, with evolution of carbonic acid. The new acid does not soften until 220°-230° C., and melts irregularly at 260°; but even if the temperature is maintained for some time at 280° C., there is no strong evolution of gas perceptible. At higher temperatures the acid sublimes, entirely without decomposition, in beautiful orange-yellow needles, which exhibit the same peculiarity in melting, dissolve in dilute ammonia, from which they are re-precipitated by acids.

If oxidised with chromic and glacial acetic acids, the acid of Graebe and Liebermann yields anthraquinon; but the new acid produces mere traces of this substance, and forms principally anthraquinoncarbonic acid. An excess of oxidising agents, however, strongly attacks the anthraquinocarbonic acid, and reduces the yield. The oxidation of anthracencarbonic acid takes place according to the equa

tion

C,H,CO,H+O,=CH,O,CO,H+H,O.

The above properties of the two isomeric anthracencarbonic acids are most readily explained, if we assume that the

acid formed by means of chlorocarbonic oxide has its carboxyl group at a, whilst that obtained from anthracensulphuric acid has the same group at b or c. If an acid of the former constitution is oxidised, carbonic acid must be evolved, because the oxygen of the quinon passes to the carbon atoms a: in the latter the oxidation of the carbon atoms at a can be completed without displacing the groups at b or c.

Methylanthracen, C15H12.

This compound was obtained by Weiler and Fischer on passing dimethylphenylmethan or dimethylphenylethan through ignited tubes filled with fragments of pumicestone. The reaction takes place according to the following equation

C15H16=2H2+C15H12.

In order to maintain and purify the methylanthracen, the coloured and adhesive matter found in the receiver is first washed with a little ether and alcohol, in order to purify it from any undecomposed residue of ditolylmethan, and then either recrystallised from glacial acetic acid and alcohol, or by preference converted into finely crystalline picric acid compound, which in turn is decomposed with ammonia.

Its melting-point varies from 198°-201° C., according to the method of purification employed.

Methylanthracen is insoluble in water, sparingly soluble in ether, alcohol, glacial acetic acid, and is deposited from solutions saturated at high temperatures in beautiful white shining scales, tinted with a slight yellowish shade. It dissolves readily in bisulphide of carbon, chloroform, and benzol, but very sparingly in methylic alcohol, aceton, and petroleum. It is precipitated from its solutions on the addition of water in white flocks, which, when dry, become gradually yellow on prolonged exposure to the air. Concentrated or fuming

nitric acid dissolves methylanthracen slowly in the cold, but more rapidly at a moderate heat. The action of concentrated or fuming sulphuric acid is similar.

Bromine, dissolved in disulphide of carbon, has a very violent action upon this hydrocarbon, dissolved in the same medium; torrents of hydrobromic acid escape, whilst finely crystalline yellow bromine compounds are formed.

With picric acid it yields a compound crystallising in fine, long, dark red needles, which very closely resemble anthracenpicric acid, and are decomposed by water or alcohol, but more readily by aqueous alkalis and ammonia.

Methylanthracen sublimes in large beautiful scales, which, when perfectly white, display a fine blue fluorescence.

If methylanthracen is oxidised with glacial acetic acid and chromic acid, we obtain not methylanthraquinon, but anthraquinon-carbonic acid, melting at 283°, which yields anthraquinon on heating with soda-lime. If methyl-anthracen is dissolved in excess of alcohol, and if ordinary concentrated nitric acid is afterwards added, a brisk reaction is set up, owing to the oxidation of the alcohol. The alcohol is allowed to evaporate spontaneously down to about onethird, and a little water is then added. On cooling, a yellow crystalline matter is deposited, which is well washed with water. The filtered solution is mixed with water, in order to obtain the residue of the dissolved body. This substance is methylanthraquinon, and is best purified by sublimation, when it takes the form of thin ramified needles or leaflets.

Dibrommethylanthracen, C15H10Br2.

If methylanthracen in disulphide of carbon is treated with two molecules of bromine, there escape torrents of hydrobromic acid; and we obtain, when the reaction is completed, and the disulphide of carbon is expelled, a bromine compound, which crystallises from chloroform, glacial acetic acid, &c.,

in fine golden yellow needles, which sublime in large thin needles, and melt at 156°.

Dimethylanthracen, C16H14.

This compound was first obtained by Van Dorp, who was led to it during the examination of a tinctorial substance prepared from cochineal. If carmine dissolved in sulphuric acid is heated at 120° C., the yellowish-red colour of the liquid becomes violet, whilst CO2 and SO, are evolved. After the temperature has been maintained for a time at 140°-150°, the solution is poured into water, and the new colour is obtained as a brown flocculent precipitate. After being washed and dried, it is extracted with boiling alcohol, and the alcoholic solution is evaporated. This compound, which was named ruficoccin by Liebermann and Van Dorp, is sparingly soluble in cold water, a characteristic by which it is distinguished from carmine and from carminic acid. It dissolves in alcohol with a beautiful yellow fluorescence, sublimes partially in red vapours, which condense in yellowish-red needles, and with mordants dyes shades resembling those of cochineal, although less brilliant. Its analysis leads to the formula C16H12O6*

8

From the origin and behaviour of ruficoccin, which approximate very closely to those of rufigallic acid and the colouring matter of opianic acid, C1HO (deducible from anthracen), Liebermann and Van Dorp concluded that it must contain the complex of dimethylanthracen, and have the constitution

(CH3)2 C11H2 (OH)4

2°

Its origin is readily intelligible, if we assume that the substance which on nitration induces the formation of nitro