182 SCHUNCK AND MBRCHLEWSKI XIX. -ATotes on Mudder Colou&ng Mattew. By EDWARD SCHUSCK, Ph.D., F.R.S., and LEOX MARCHLEWSKI, Ph.D. I. CONSTITUTION O F RVBIADIN. Irr our prexrious communication (Trans., 1893, 63, 969), we gave an account of the properties of this substance, and arrived at the conclusion that it must be considered as a fiornologue of purpnyo- xanthin, which it closely resembles. As formerly mentioned (Eoc. cit., p. 1137), experiments were made with a view t,o determine the constitution of rubiadin, the chief point being to ascertain the position of the methyl group in the molecule. These showed that the methylpurpuroxanthin [CH, : OH : OH = 3’ : 1 : 31 obtained, though very similar to, is not identical with, rubiadin, as its melting point is considerably lower than that of the latter.We are now able to confirm the views formerly expressed as regards the nature of rnbiadin, more especially as to the character of the methyl complex contained in it. The analyses given in our first memoir show that the composition of the substance is undoubtedly that of a dihydroxyanthraqninone containing an additional CH,. Now this CH2 might either be a constituent of a methoxy-group, or it might be a constituent of a methyl group united to a carbon atomON MADDER COLOURINGF MATTERS. 183 of the anthraquinone complex. In order to decide between these two views, we treated rubiadin with hydriodic acid by Zeisel’s process, and found that it did not yield methylic iodide, and could not, there- fore, contain a methoxy-group. This conclusion was confirmed by the following experiment.A small quantity of rubiadin was heated with concentrated sulphuric acid for 15 minutes at 180” ; the solution, when cold, was poured into water, the product extracted with ether, and the ethereal solution evaporated. The residue, when dissolved in benzene, gave crystals of unchanged rubiadin, which melted at about 290”. These experiments proved the absence of any methoxy-group, and it now remained to determine the position of the methyl group in the molecuie. On oxidation, rubiadin yields phthalic acid, so that the methyl group must be contained in the same nucleus as the hydroxyl groups. The experiment was conducted as follows. Rubiadin was boiled with a mixture of glacial acetic acid and chromic acid, and the solution, when cold, was poured into water, and extracted with ether; the ethereal extract, on evaporation, left a greenish mass, which was again extracted with ether, and the filtered solution again evaporated.In this way a small quantity of a colourless substance was obtained, which sublimed easily. The sublimate melted at 128”, and on being cautiously heated with resorcinol, and the product dissolved in caustic soda, gave the characteristic green fluorescence of fluoresceln, proving that it was pht,halic acid. Rubiadin must, therefore, have one of the two following formulae. CO OH CO OH TVhich of these two is the true one could only be decided by a sgnthetical experiment. I n order to obtain a methylpurpuroxanthin having the constitution [OH : CH, : OH = 1 : 2 : 31, we employed benzoic acid and metadihydroxyparatoluic acid.The condensation of these t x o substances in presence of sulphuric acid would take place in accordance with the following equation OH CO OH The dihydroxyparatoluic acid required was prepared by ourselves f ~ o m Kahlbaum’s pamtoluic acid ; its melting point was found to be lis”, as given by Weinreich (Ber., 20, 98.2). The condensation184 YCHUNCK AND MARCHLEWSKI was effected in the manner described in a previous paper by one of us (Zoc. cit., p. 1142), but the heating was continued longer, as i t was found that condensation took place with some difficulty. A mix- ture of 4 grams of dihydroxypaiatolnic acid, 15 grams of benzoic acid, and 200 grams of snlphuric acid was heated for 15 hours at 110-120".The solution was then poured into water and shaken with ether ; the ethereal extract was evaporated, and, the residue having been sus- pended in water, the excess of benzoic acid present was driven off by means of a current of steam. This was~followed by treatment with boiling benzene, and the solution, on cooling, deposited crystalline needles, which, after recrystallisation three times from hot benzene, yielded the following results on analysis. 0.1201 gave 0.3113 CO, and 0.0448 H,O. C15HloOa requires C = 70.86; H = 3.93 per cent. At 270", the product begins to sublime, and melts at about 282", but if heated qnicklyit melts at 290°, which is the melting point of rubiadin itself. The solution in concentrated sulphuric acid is brownish-red, and shows a narrow absorption band in the red, which is a little nearer the less refrangible end of the spectrum than the correspondirg band of a purpuroxanthin solution The substance is easily soluble in alcohol or ether, and is obtained in orange-coloured needles on evaporating the ethereal solution.I t s properties do not therefore differ greatly from those of rubiadin, but seeing that there was some uncertainty as to its precise melting point, an uncertainty due to its beginning to sublime before complete fusion, we determined to find some other means of com- parison between the natural and the artificial products, and for this purpose had recourse to the acetyl derivatives. C = 70.69 ; H = 4.13. It dissolves in alkali, forming a red liquid. Acet y lrubiadiu. I n order to acetylate rubiadin i t was digested with acetic anhydride and it small fragment of stannous chloride.The excess of acetic anhydride was decomposed with alcohol, and after evaporating the ethylic acetate, the residue was crystallised from alcohol. In this way a product was obtained crystallising in lustrous, silky needles, and melting at 225" ; it is not changed by cold alkalis, but is decom- posed on heating with them. Acetyl Compound of the Nethylpurpuroxanthin [OAc : CH, : OAc = 1 : 2 : 31. This compound was prepared in the same way as the preceding. After two crystallisations from alcohol, it melted at 217-218", and differed in appearance from acetylrubiadin, the latter crystallising inON MADDER COLOURING MATTERS. 185 fine needles, whilst the former appears in crystalline p i n s .Hence it follows that the second of the two formulse given above is the one belonging to rnbiadin. This result is of some interest, for as pur- puroxanthincarboxylic acid is held t o be a derivative of a-anthra- quinonecarboxylic acid, and may therefore have the following formula, CO OH it is possible that rubiadin may be the parent substance of this acid, which would therefore be formed from rnbiadin by some process of oxidation going on in the plant. JI. THE MONALKYL ETHERS OF ALIZARIN. Several years ago a monomethylalizarin and a monethylalizarin mere prepared by one of us by heating alizarin with methylic iodide or ethylic iodide and caustic potash, in sealed tubes (Munchester Memoirs, 1873). During this process, strange to say, no dimethylalizarin or di- ethylalizarin is formed, from which it follows that the etherification of one of the hydroxyls of alizarin is not easily effected.This behaviour is not singular. In their memoir on the derivatives of monhydroxy- xanthone (Bey., 1893, 76), Kostanecki and Dreher stat'e that whilst the 2-, 3-, and 4-hydroxyxanthones may be easily methylated, l-hydr- oxyxanthone, in which the OH group adjoins the carbonyl group, remains unchanged when treated with potash and methylic iodide. The same may take place in the case of alizarin, in which the hydr- oxyl group adjoining the carbonyl group is probably the one which is not easily affected. It would appear therefore that the constitution of monomethylalizarin is represented by the following formula.CO OH It melts at 228-229", and may be easily acetylated by Lieber- The acetyl compound crystallises from alcohol in It yielded, on analysis, C = 68-57 ; H = 4.21. mann's method. light yellow needles, and melts at 186-187". the following result's. 0.1187 gave 0.2984 CO, and 0.0450 H20. C,,H,O2(0CH3).O*C2H3O requires C = 68.92 ; H = 4-03 per cent. We endeavoured also to prepare the second possible methylalizarin186 ON MADDER GOLOURING MATTERS. from ruberythric acid. If we assume, with Liebermann and Bergami, that ruberythric acid contains one of the two hydroxyls of alizarin in a free state, and assuming, further, that the union of the alizarin with the sugar takes place through that hydroxyl which may be easily alkylised, it might be possible to obtain a methylrnberythric acid, and, by decomposing the latter, to arrive a t the second methyl- alizarin sought for.Our experiments failed; we were unable to obtain a, methylruberytbric acid by treating the potassium salt with methylic iodide and methylic alcohol. It is, of course, quite possible t h a t , under certain conditions, especially such as occur in t,he rege- table organism, a compound of this kind may be formed, yielding, as a product of decomposition, the monomethylalizarin of Hiimmel and Perkin (Trans., 1893, 63, 1174). BcetyZcfhyInZizarin.-E thylalizarin melts at 188-189", and may be easily acetylised ; the product crystallises from alcohol in yellow needles, and melts a t 141". Analysis yielded the following results. 0.1130 gave 0.2878 C02 and 0.0482 H,O.C = 69.46; H = 4-73. C1PH~OZ(OC2Hj)*0*CzH30 requires C = 69.67 ; H = 4.51 per cent. 111. RCBERYTHRIC A C I D . In connection with previous work on the glucosides (Zoc. cit., p. 1137), which led to the conclusion that the constitution of these compounds is in harmony with Tollens's glucose formula, we examined the action of phenylhydrazine on the glucoside of alizarin. After several ex- periments, we came to the conclusion that the two substances do not react, a fact which points to the absence of aldehyde groups in ruberythric acid. The formula of this glucoside must, therefore, also be in accordance with Tollens's glucose formula, or a similar one; it may, at the same time, be mentioned that, according to Liebermann and Bergami, the sugar is contained in this compound as a biose residue.Ruberythric acid, according to Liebermann and Bergami, yields an octacetyl derivative. We have succeeded in obtaining benzoyl derivatives also, using €or this purpose the method adopted by Baum, Banmann, and Schotten. On shaking up a solution of ruberythric acid in caustic soda (1 part of soda to 8 of water) with benzoyl chloride, a yellow precipitate was formed, which, after being collected, washed first with dilute caustic soda and then with hot qater, was dried and dissolved in benzene. On partial evaporation of the benzene, the benzoyl derivative separated in dotty masses, which appeared amorphous, even under the microscope. Its composition was that of a heptabenzoylruberythric acid, as the following analysis shows.INTERACTION OF BENZYLAMINE, ETC. 187 0.1521 gave 0.3863 C02 and 0.0658 H20. C,5H,02, requires C = 69.65; H = 433 per cent. The compound is insoluble in dilute alkali, but, on boiling, it gradually dissolves, and the solution then contains alizarin, benzoic acid, and sugar, It is decomposed by hot dilute sulphuric acid, yielding the same products. A quantitative determination of these products of decompasition was not thought necessary, as the number of benzoyl groups was sufficiently indicated by analysis. I f a weaker solution of alkali (1 part NaHO to 10 parts H20) be used, a hexa- benzoyl derivative is obtained, as shown by the following analysis. C = 69.29; H = 4.81. 0.1349 gave 0.3396 CO, and 0.0.529 H,O. C6,H,,0zo requires C = 68.60; H = 4.33 per cent. Rubery-thric acid, therefore, behares like glucose as regards the benzoyl derivatives which it yields, different products being obtained according to the concentration of the soda lye emploFed (v. Skranp, Xufiats. Chem., 10, 389, and Baumann, Bey., 19, 3219). C = 68.65 ; H = 4.36