1566 PERKIN AND HUMMEL : THE COLOURIWO MATTERS CII1.-The Colotwiq Matters occurrhg ill cccrioiis Bj-itish Plctnts. Part I. By ARTHUR GEORGE PERKIX and JOHN JAMES HUMMEL. SIWE the introduction of coal-tar colours, comparatively little atten- tion has been paid to the colouring matters of plants, especially of British plants ; it is well known, however, that, in the past, many of these have been employed in dyeing, and that this is the case even at the present time in remote districts, notably in the Highlands of Scotland, some few native plants being used for this purpose. By means of an extended series of dyeing experiments,* it has been ascertained by one of us that many British plants appear to be sufficiently rich in colouring matter (chiefly Yellow) to make them worthy of chemical examination.Jn many cases this will probably be a matter of some difficulty, partly because it is not always easy to obtain the requisite quantity of raw material, and partly because the amount of colouring matter present, even in the best circumstances, will be only very limited. The study of the various natural yellow colouring matters is particularly interesting, however, at the present time, because their chemical constitution, of which chemists have long been ignorant, is now being gradually made known, and it may be that Eome of those occuri*ing in British plants will fill up gaps in the comparatively limited series with which ve are at present acquainted. Even in the event of many proving to be identical with those already known, the investigation sliould still have an interest as showing the wide distribution of certain colourirlg matters hitherto only known to exist in a few plants.The Colowing Matters contaitied iir the Yellow IVaZZ$ou~er (Cheirantlins ckeiri). A preliminary dyeing experiment with the purplish-brown petals of the common garden wallflower showed them to be comparati~elg rich in colouring matter, but since the colour with alumina \xis not * The results of these dyeing experiments will be communicated ekewhcrc at a later date.OCCURRING IN VARIOUS BRITISH PLANTS. 1567 pure, being greenish-olive yellow, a similar experiment was made with the bright yellow Aowers of the variety known as " Cloth of Gold." As expected, these yielded a bright yellow with alumina mordant, and, indeed, their general dyeing properties were very similar to those of such well-known dye-stuffs as quercitron bark, &c, For t-he purpose of chemical examination, a pound of the flowers was extracted with boiling water, and after allowing the decoction to cool, and filtering off deposited flocculent matter, the filtrate was acidified with sulphuric acid, and boiled.On cooling, a moderate quantity of an olive-yellow precipitate separated ; this was collected, washed, dried, and examined as follows. The semi-crystalline product was digested with boiling alcohol, when the colouring matter passed into solution and a nearly colour- less, crystalline residue of calcium sulphate was left. On evapo- ration, the alcoholic extract deposited crystals, but these were intermingled with a wax-like substance, which could not be readily removed by recrystallisation ; the concentrated solution was, therefore, poured into a large volume of ether, the mixture washed several limes with water, and then extracted with dilute alkali, which removed the colouring matter, leaving the wax in solution in the ether. The alkaline liquid was neutralised, the precipitated colouring matter again dissolved in ether, and the ethereal solution evaporated to dryness.On examination, the bright yellow product was found to consist of two substances, for, when suspended in boil- ing acetic acid, and treated with hjdrobromic acid, an orange- coloured, crystalline hydrobromide was produced, but intermingled with particles of a yellow substance, which refused to react with the Iialoi'd acid in this way.Experiment, moreover, showed that these two Substances were readily distinguished from one another by t h e difference in their solubilities in alcohol ; this was sufficiently marked to render their separation comparatively easy. As the more soluble colouring matter was present in by far the greater proportion, this was first examined. It was best purified by crystallisation from alcohol and water, and was thus obtained in the form of glistening, yellow needles. 0.1362 gave 0.2549 CO, and 0.0325 H,O. CI5Hl007 reqnires C = 59.60; H = 3-31 per cent. Dilute alkalis dissolved it with a yellow coloration ; with lead acetate in alcoholic solution, an orange-red precipitate was formed, whilst alcoholic ferric chloride gave a dark green coloration.Treated with mineral acids in the presence of acetic acid, orange to orange-red crystalline compounds were produced, SO that it is evident that not this but the mwe insoluble colouring matter was uureactive with C = 59.82; H = 3.10.1568 PEREIN AND ETUMMITL : TIIE COLOURIX'GI MATTERS the haloid acids. Digested with acetic anhydride and sodinnl acetate in the usual way, an acetyl compound was obtained, crjstal- lising from alcohol in colourless needles melting a t 189--19lo. 0 1271 gave 0.2730 CO, and 0.0477 H,O. C = .W58 ; H = 4.16. C,,H5Oi(C2H,O), requires C = 58.59 ; H = 3-90 per cent. The colourhg matter was digested with fused alkali a t 180-200° for half an hour, the melt dissolved in water, neutralised with acid, and the products of the action extracted from the solution with ether.The small quantity of crystalline matter thus obtained was found to be a mixture of two substances which could be readily seprrrated, in that, in aqueous solution, one only gave a precipitate with lead acetate; on decomposing this, it yielded a substance crjstallising from water in colourless needles melting a t 195O, and identical with protocntechuic acid. The filtrate from the lead precipitate gave the yhloroglucinol reac- tion. There coiild be no doubt, therefore, that tjhe more soluble colonring matter, C15H1007, is puercetin, a fact which was further corroborated by its dyeing properties being identical wit'h those of quercetin pre- Fared from quercitron bark. Spaiingly Soluble Colouring Matfey.--This was purified by several crjstallisations from alcohol, and was thus obtained in the form of minute, yellow needles.0.1141 gave 0.2533 C 0 2 and 0.0420 H,O. C,,H,,O, requires C = 60.75 ; H = 3.79 per cent. Although considerably less soluble in alcohol and acetic acid than quercetin, yet, in its principal reactions, it closely resembled it, for it dissolved in dilute alkaline solutions with a yellow colour ; with lead acetate, it gare an orange-red precipitate ; and with alcoholic ferric chloride a dark green coloration. As before stated, when treated with halo'id acids in the presence of boiling acetic acid, it did not yield a compound with the acid, although, with sulphuric acid in a similar manner, an orange-rcd, crystalline substance was produced. Its behaviour in this respect corresponded with that of the known methyl ethers of quercetin, rhnmnetin (quercetin mononiethyl ether), rhamnazin (quercetin dimetl yl etl.er), a a d quercetin tetramethyl ether, which have been previously studied by one of us and L.Pate ('rrans., 1895, 650 ; 1F96, 1443), and it was, therefore, suspected that this colouring matter also contained a methoxg-group. An experi- ment by Zeisel's method gare the following result: C = 60.54; H = 4.09. 0.1437 gave 0.1040 AgI. CH, = 4.62. C15H,0,*OCH, requires CH, = 4.68 per cent.OCCURRING IN VARIOUS BRITISH PLINTS. 1569 It, therefore, contained one methoxy-grodp. To the hydriodic acid residue, after dilution with water, sodium hydrogen sulphite solution was added, and the yellow, floxulent product collected arid crysta!- lised from dilute alcohol ; it was thus obtained as a glistening mass of yellow needles, and appeared to be pwsrcstin. This wag coqfirmed by converting it into an acetyl corupound; the latter crystallised from alcohol in colourless needles melting a t 189-191', and was found to be ideiiticrtl with acetylquercelin.Tho colouring matter, C16H1?07, is, therefore, a quercetin monomethyl ether. The only known monomethyl ether of quercetin is rhamnetin, present in Pers:an berries i n the form of a glucoside, xanthorham- nin. As, in appearance and ganeral properties, the colouring matter, C16H,20,, was very similar to rhamnetin, it seemed possible fhat the two were identical ; to determine this point, the small quantity of the substance, ClsH1207, that remained was converted iuto an acetjl compound, and this crystallised in colourless needles melting at 195-196'.It, therefore, appeared that this siibstance could not be rhamnetin, for the melting point of its acetyl compound is given by Liebermann and Hormnnn ( B e y . , 11, 1618) as 181--18;3O, and by Herzig (Monatsh., lW3, 9, 548) as 183-185O. Since the publica- tion of these papers, however, it has been shown by one of us and J. Geldard (Trans., 1895, 67, 496), that Persian berries contain, not only rhamnctin and querceth, but a third substance, rhamnaziri, and it was thus possible that the melting point of acetylrhamnetin might be higher than that given above. Experiments, however, with rhnmxietin, which had been submitted to puriticstion in various ways, failed to support this supposition, the acetyl compound in each case melting at 185-186'.Moreover, when the acetyl compound cf rhamnetin and that of the substance Cl6H;,O, were crystallised from alcohol side by side, and under similar circumstances, they could be readily clistinguiBhrd from one another; for, whereas the latter always separated as a semi-solid, spongy mass of hair-like needles, the crystals of the former, which were considerably larger in size, readily sattled down in the mother liquor on agitation. There could be no doubt, therefore, that the substance, C16H1207, is a new quercetin monomethyl ether, and for it we propose the name Isorhamnetin. As it is not possible to obtain a further supply of the flowers Lill next summer, some time must elapse before attempting to decide the position of the methoxy-group in this substance. Eren f,hen the small amount of this product which is present in the plant, must recder this a work of considerable difficulty, and may delay the results shill fprther.The t o t d quantify of isorhamnetin obfained from tho flowers being not mom than 0.4 gram, sufficient mas not available f o r any thorough1570 PERKIN AND HUMMEL: THE COLOURXNO MATTERS dyeing experiments. A single trial, using the ordinary striped mop- danted calico, indicated that its properties, i n this respect, were, also, closely similar to those of both quercetin and rhamnetin-these two colouring matters themselves yielding in this way almost identical shades. Experiments in this direction will b3 carried out in detail when more of the colouring matter is to hand.The Colouring Matter in White Hawthom Blossom (Cratagus Oxyacantha) . Examination having showed that white hawthorn flowers contain it yellow colouring matter, and, the present season being particularly favourable, about 20 lbs. of the flowers were collected for examination. These were treated in the manner above described in connection with the examination of wall-flowers. In this case the precipitate obtained was of a rich chocolate-brown, and, although produced in moderate quantity (as will be seen below) it yielded, ultimately, a very small quantity of pure colouring matter. The isolation of the pure colouring matter from this product, by t.he methods usually found serriceable for such a purpose, gave exceedingly unsatisfactory results, as, in each instance, it was so contaminated with resinous matter, that its purification could only be accomplished with con- siderable loss.Eventually a method, similar to that previously employed for the isolation of the colouring matter of Quebracho coloyado (Trans., 1896, 69, 1303), was found to be the most service- able. For this purpofie, the crude product was dissolved in alcohol, poured into a large quantity of cold water, excess of sulphnric acid added, and the mixture heated to the boiling point. As the tempera- ture rose, the suspended, brown, flocculent precipitate gradually aggregated, forming a black, tarry mass, and the digestion was con- tinued until the supernatant liquor, at first somewhat milky, became clear; this, after decantation, mas extracted with ether, and the extract evaporated, a light brown, sticky product being thus obtained, from which crystals of the colouring matter separated after long standing.Digestion with boiling chloroform remored from this some of the impurity, these extracts being placed aside for examin% fion (A). The residue was now further purified by. crystallisation from dilute alcohol, but still contained a brown matter which could not be removed by similar treatment ; i t was, therefore, converted into an acetyl derivative. This, which crystallised from alcohol in colour- less needles, was decomposed in the usual way, and the regenerated colouring mattel* crystallised from dilute alcohol. By this means only 0.5 gram of the purified product was obtained, but i t is most probable that a portion was retained by the tarry matter mentionedOCCURRING IN VARIOUS BRITISH PLANTS.1571 above. It was subsequently found that, by treating an aqueous decoction of the flowers with lead acetate solution, decomposing the precipitated lead compound with sulphuric acid, and extracting the acid solution with ether, the colouring matter could be obtained in a condition which lent itself more readily to purification. At this time, however, no more raw material was available, though, fortu- natcly, the quantity of substance obtained by the former method was found to be sufficient for the identification of this colouring matter. C = 59.40 ; H = 3.61. 0.1100 gave 0.2396 CO, and 0.0358 H,O. C1,H,oO, requires C = 59 60; H = 3.31 per cent.It formed a glistening mass of yellow needles, readily soluble in alcohol, mid soluble in alkalis with a yellow coloration. In alcohoIic solution, lead acetate yielded an orange-red precipitate and ferric chloride a dark green colorcdion. With mineral acids, it yieIded crjs- talline compounds. AcetyZ Derivutiz;e.--By crystallisation from alcohol this mas obtained 8 6 colonrless needles melting at 189-191'. 0.1195 gave 0.2556 CO, and 0.0437 H20. C = 58-33; H = 4.06. Action of Fused AEkaZi.-For this purpose, but 0-1 gram WBS arail- able. The products of the action were readily separated by means af lead acetate in aqueous solution, in that ouly one of them yielded a, precipitate with this reagent. By decomposition, i n the usual way, the lead compound yielded a trace of a crystalline substance, a solu- tion of which gave with ferric chloride a green coloration, from whicli it appeared to be protocatechuic acid.The filtrate from the lead pre- cipitate gave the phloroglwcinol reaction. As was to be expected from the above results, a dyeing trial corroborated the fact that the colouring matter of hawthorn blossom i s quercetin. On nllo.cviiig the chloroform extract (A), obtained during the puri- fication of the colouring matter, to evaporate spontaneously, it deposited a very small quantity of crystalline substance intermingled with a brown sticky product. B y recrystallisation, first from benzene and then from water, it was obtained in colourless needles melting at 177-17&', and subliming unchanged, soluble in dilute alkalis, but yielding no coloration with ferric chloride in aqueous solution. The quantity available was far too small for analysis, and its examina- tion must, thereEore, await a further supply of raw material, Its general properties, however, suggesteii that it might be veratric (dimethylprotocatechuic) acid. It will be of interest also, if possible, to isolate and study the sub- stance which yields the brown precipitate when the extract is digested C15H507(CZH30)5 requires C = 58.59 ; H = 3.90 per cent.1572 CEATTAWAY: THE CONSTITUTION OF THE with boiling dilute acids. This reaction snggests that it is possibly's catechin-like compound, the brown prodact, i n its nature, resembling t h a t formed when either catechin, or cyanomaclurin is similarly treated. That the above colouring matters exist in these flowers in the form of glucosides there c m be little doubt, and at a convenient opportunity they will be subjected to examination. For the present, however, it appears to us that more interest is attaf;hed to the examination of the colouring matters themselves than to their glucosides, it being possible that in some of these plants missing members of the qnercetin or xanthone series may exist. Clothworkers' Research Laboratory, Dyeing Department, T'orkshire Col Eege.