THE CONSTITUENTS OF RED CLOVER FLOWERS. 231XXVI.-The Constituents of Red Clove?- Flowem.By FREDERICK BELDING POWER and ARTHUR HENRY SALWAY.THE flowers of the common red clover (TrifoEium pratense, LinnB)have been used to some extent medicinally in recent years onaccount of their assumed alterative properties, and have even beenrecommended in the treatment of cancer (Anter. J . Pharm., 1881,53, 85). So far as known to us, these flowers have never beensubjected to a chemical examination, and the only knowledge oftheir constituents appears to be embodied in a statement that theycontain tannin, two resins, fat, and chlorophyll. I n this connexionit may incidentally be noted that Perkin and Phipps, in a paperentitled “ Notes on some Natural Colouring Matters ” (Trans.,1904, 85, SS), have recorded that it cursory examination of theflowers of the white clover (Trifolium repens) indicated the presenceof quercetin in the form of a glucoside.The same authors further-more remark that (( this colouring matter was recognised by themelting point of its acetyl derivative, and its decomposition productswith caustic alkali, and in consequence of these observations afuller investigation appeared unnecessary.”In view of the above considerations, and the fact that red cloverflowers are available in practically unlimited quantities, it appeareddesirable that a complete study of their constituents should beundertaken. The present investigation has disclosed the presenceof slr considerable number of new and interesting substances, anda summary of the results obtained is given at the end of thispaper.EXPERIMENTAL.The material employed in this investigation was carefullycollected for us under the superintendence of Mr.P. E. F. PerrBdBs,B.Sc., F.L.S., during the month of June, from a, field of cultivatedred clover (Trifolium prutense, LinnB) in Kent. The flowers orblossoms were separately gathered, and were therefore, so far aspossible, free from the green, herbaceous parts of the plant232 POWER AND SAJ,WAY:A portion (10 grams) of the dried flowers was tested for analkaloid, but the reactions obtained were so slight as to indicatcthe presence of not more than traces of such a substance.A further portion (25 grams) of the dried flowers was successivelyextracted in a Soxhlet apparatus with various solvents, when thefollowing amounts of extract, dried in a water-oven, were obtained :Petroleum (b.p. 35-50") extracted 0.50 gram 1 2.0 per cmit.Ether ,, 0.50 ), = 2.0 ,,Cliloroform ,, 0.35 ,) = 1'4 ,,Ethyl Acetate ,* 1.50 ,, = 6'0 ,,Alcohol ,, 4'50 ,, = 18.0 ,, - -Total 7.35 grams = 29-4 per cent.For the purpose of a complete examination, 264 kilograms of theflowers were collected. This material, after careful drying,amounted to 58.5 kilograms, or 22.16 per cent. of the originalweight. The. entire amount of this material was coarsely powdered,and then extracted by continuous percolation with hot alcohol.After the removal of the greater portion of the alcohol, which wasconducted a t a temperature not exceeding 8 5 O , a dark green, viscidextract was obtained, which amounted to 15.78 kilograms.7)istiklntio.n of t h Rafmrt with Steam..Separation, of an, Essrn tintOil.A quantity (1.5 kilograms) of the above-mentioned extract wasbrought into a suitable apparatus with a little water, and steampassed through the mixture for several hours. The distillate, whichcontained some oily drops, was extracted with ether, the etherealliquid being washed, dried, and the solvent removed. A smallamount (1.5 grams) of an essential oil was thus obtained, whichpossessed a rather unpIeasant odour. On subsequently treating10 kilograms of the alcoholic extract in the manner above described,a further quantity (10.5 grams) of essential oil was obtained.Theyield of oil was thus equivalent to 0.028 per cent. of the dried, or0.006 per cent. of the fresh clover flowers. This essential oil, whendistilled under diminished pressure, passed over between 90° and160°/20 mm. as a light yellow liquid, which, however, on keeping,ultimately became reddish-brown, and was found to contain fnrfur-aldehyde. It possessed the following constants : (I! 2Oo/2O0 - 0.9476 ;an +4OOf in a l-dcm. tube.iVon~-i~oZntiZa Constitmmts of t h e Elrf rnct.After the removal of the volatile constituents of the extract, bydistillation with steam, as above described, therc remained in tliTHE CONSTITUENTS OF RED CLOVER FLOWERS. 233(listillation flask a reddish-brown, aqueous liquid ( A ) and a quantityof a viscid resin (R), which became solid on cooling.The resin wascollected, and thoroughly washed with water, the washings beingadded to the main portion of the aqueous liquid.Examination of the Aqueous Liquid (A).The aqueous liquid was concentrated under diminished pressure,and then repeatedly extracted with large volumes of ether. Theethereal extracts, which were of a light green colour, were united,washed with a little water, dried, and the ether removed, when10 grams of a dark-coloured, soft solid were obtained. This wasfound by a preliminary examination to consist of a complex mixtureof crystalline compounds, and, in order to obtain a quantitysufficient for their separation and characterisation, 10 kilograms ofclover extract were treated as above described, the resulting aqueousliquid being extracted with ether.The yield of ethereal extractfrom this quantity of material was 64 grams.Isolation of a Xew Phenolic Substance, Pratol,C,,H80,(OH)*O* CH,.The above-mentioned ethereal extract was digested with aquantity of ether insufficient to dissolve the whole, and the sparinglysoluble portion, which amounted to 3 grams, separated by filtration.This product was repeatedly crystallised from alcohol, when asubstance was obtained which separated in colourless needles,melting a t 253O:0.1265 gave 0.3307 C02 and 0.0513 H,O. C=71.3; R z 4 . 5 .0.1213 ,, 0.3161 CO, ,, 0*0500 H2O. C=71*1; H=4.6.C,,H1204 requires C = 71.6 ; H = 4.5 per cent.The above compound, when heated with acetic anhydride, yieldedan acetyl derivative, which separated from alcohol in featheryneedles, melting a t 1 6 6 O .This was analysed, and its molecularweight determined, with the following results :0-1039 gave 0.2646 CO, and 0.0420 H,O.0.1485, in 33.26 of acetic acid, gave A t - 0'057O.Cl6Hl1O4(CO*CH3) requires C = 69.7 ; H =4.5 per cent. M.W. = 310.The substance C16H1204 was found to contain one methoxylgroup, as determined by Perkin's modification of the Zeisel method :0.0995 gave 0.0785 AgI. MeO=10*4.C1,€I,O3(0Me) requires Me0 = 11.6 per cent.No compound of the formula Cl6HI2O4, pcssessing the sameC = 69.5 ; H=4.5.0.1069 ,, 0.2715 CO2 ,, 0.0436 H,O. C=69*3; H=4*5.M.W. = 305234 POWER AND SALWAY:properties as that above described, appears to have hitherto beenrecorded.It is therefore proposed to designate the substance fromred clover flowers as pratol, with reference to its phenolic characterand the specific botanical name of the plant.Pmtol, C,,H80,(OH)*O*CH,, crystallises in a very characteristicform, since it separates from alcohol in needles, which, under themicroscope, are seen to have a talon-like shape with curved edges.It is moderately soluble in hot alcohol, but only sparingly solublein water, ether, chloroform, or benzene. It dissolves readily inhot aqueous sodium carbonate and sodium hydroxide, yielding paleyellow solutions. When dissolved in acetic anhydride, and a dropof sulphuric acid added, a yellow coloration is produced. Withferric chloride no appreciable change of colour was observed.Pratol is isomeric with several flavone derivatives, such as the2-met hox y- and 3-me thoxy-flavonol,prepared by Kostanecki and his co-workers (Ber., 1904, 37, 775;1905, 38, 993).The general behaviour of pratol is also very similarto that of the above-mentioned substances, and it therefore seemsprobable that it represents one of the many hydroxymethoxy-flavones which are theoretically possible.Acetylpratol, Cl,Hl,0,(CO*CH3), prepared, as above noted, byheating pratol with acetic anhydride, is very soluble in hot alcohol,but only moderately so in chloroform or benzene.The ethereal liquid from which the sparingly soluble pratol hadbeen separated by filtration, as above described, was subsequentlyshaken with an aqueous solution of ammonium carbonate (a) untilnothing further was removed by this reagent, then with successivesmall portions of aqueous sodium carbonate ( b ) , and finally withaqueous sodium hydroxide (c). These extracts were then separatelyexamined.Isolation of Scclicylic Acid.The liquids obtained by extraction with ammonium carbonate ( a )were united, acidified with sulphuric acid, and extracted withether.This ethereal liquid, on the removal of the solvent, yieldeda quantity (20 grams) of a dark green oil. The latter was digestedwith considerable quantities of hot light petroleum, these liquidsbeing decanted, and the solvent evaporated. I n this manner2 grams of colourless needles were obtained, which, after a fewcrystallisations from water, melted a t 1 5 4 O and gave an intenseviolet coloration with ferric chloride THE CONSTITUENTS OF RED CTAOVER FLOWERS.2350.1230 gave 0.2765 COz and 0.0507 H,O.This substance was thus identified as salicylic acid.The portion of the ammonium carbonate extract which wasinsoluble in light petroleum was heated for some time with methylalcohol in the presence of concentrated sulphuric acid. A productwas thus obtained which could be further separated into phenolicand non-phenolic methyl esters by shaking its ethereal solutionwith dilute aqueous sodium hydroxide. The portion insoluble inthe alkaline liquid, consisting of the non-phenolic esters, amountedto 10 grams, and was a dark-coloured oil. It was hydrolysed withalcoholic potassium hydroxide, and the regenerated acids distilledunder diminished pressure, when fractions were collected a t100-130° and 130-230°/20 mm., whilst a quantity of undistillableresin remained in the flask.These fractions were further examinedby neutralising them with aqueous sodium carbonate, and frac-tionally precipitating the solutions with silver nitrate. A seriesof silver salts w?s thus obtained, which, with the exception of thelast fraction (111), were more or less brown in colour. After dryingin a vacuum over sulphuric acid, they were analysed:C = 61.3 ; H = 4.6.C,H60, requires C = 60-9 ; H =4*3 per cent.Fraction 100-13Ooj20 mm.Fraction 130-230°/20 mm.0.1760 of salt gave 0.0852 Ag. Ag=48-4.(I) 0.4969 of salt gave 0-1904 Ag.Ag=38.4.(11) 0.1531 ,, ,, 0.0700 Ag. Ag=45.7.(111) 0.7246 ,, ,, 0.4444 Ag. Ag=61.3.It was evident from these results that the non-phenolic acidsconsisted of a complex mixture.Identification of p-Coumaric A cia?.The portion of the above-mentioned methylated product whichwas soluble in aqueous sodium hydroxide was benzoylated by theSchotten-Baumann method, and then subjected to fractionalcrystallisation from alcohol. After a quantity of gummy, amor-phous matter had been separated, a crystalline substance wasdeposited, which was ultimately obtained in slender needles, meltingat 129O:0.0783 gave 0.2079 CO, and 0.0390 H20.BzO*C6H,-CH:CH*C0,Me requires C = 72.3 ; H= 5.0 per cent.This compound is thus seen to agree in composition with thebenzoyl derivative of methyl p-cournarate, which hitherto appearsnot to have been described. A little of it wits therefore preparedfor the purpose of comparison, when it was found to crystallise inC = 72.4 ; H = 5.5236 POWER AND SALWAT:slender needles, melting a t 129O, and to be identical in all respectswith the compound obtained from the acid present in red cloverflowers.The identity of the latter acid was further confirmed bythe hydrolysis of the benzoyl derivative of its methyl ester, whenit substance was obtained which crystallised from water in colourlessneedles, melting at 215O with evolution of carbon dioxide, andotherwise exhibited all the characters of p-coumaric acid.Isolation of a New P ellow Compound, Cl6H1,,O7.The sodium carbonate extracts ( b ) , obtained as above described,were separately acidified, and the resulting precipitates collectedand examined.The product obtained from the first sodiumcarbonate extract consisted chiefly of chlorophyll, and nothingcrystalline could be isolated from it. The second sodium Carbonateextract yielded a yellow powder, which was dissolved in hotalcohol, in which it was freely soluble. This solution, after beingkept for some time, deposited a small amount of a, colourless,crystalline substance, which, after purification, melted a t 253O, andyielded an acetyl derivative melting a t 166O. It was found to beidentical with the substance C16H1204, designated as pratol, whichhas already been described. As the mother liquors, which stillcontained all of the yellow compound, deposited nothing furtheron standing, the alcohol was completely removed, and the residueacetylated by means of acetic anhydride.On distilling off thegreater part of the latter liquid, a crystalline product separated,which proved to be a mixture of substances. It was fractionallycrystallised from alcohol, when a small quantity (1.0 gram) of anncetyl derivative was obtained, which appeared to be homogeneous,and separated in colourless, glistening, prismatic needles, meltinga t 145-147O with evolution of gas:0.1530 gave 0.3347 CO, and 0.0555 H,O.The acetyl content of this compound was determined by heatingi t for some time with dilute sulphuric acid in the presence ofalcohol, the alcohol being then completely removed, and the pre-cipitated yellow substance, which was insoluble in cold water,collected and weighed :0.1690 gave 0.1108 of yellow substance, or 65.6 per cent.C,,H,O,Ac, requires C = 59.8 ; H = 3.7 ; C,,H,,O, = 65.1 per cent,.C,,€I,O7Ac4 ,, C = 59.5 ; H = 4.1 ; C,,H,;07 = 65.3 ,,It would appear from these results that the yellow compoundpossesses the formula CI6H,,O7, and that it' contains four hydroxylgroups.The presence of a methoxyl group was also established.The compound, C16H,UOi, as obtained from its scetyl derivative,C=59.7; H=4*0THE CONSTITUENTS OF RED CLOVER FLOWERS. 237crystallises from boiling dilute alcohol in thin, yellow plates, whichmelt and decompose at about 280O. It is readily soluble in alcoholor acetic acid, but sparingly so in water, ether, chloroform, orbenzene.It dissolves in alkalis with a yellow colour, and its solutionin concentrated sulphuric acid exhibits a brilliant greenfluorescence. In alcoholic solution it gives with ferric chloride abluish-black coloration. The amount of this substance availablcwas not sufficient for its further examination.Isolution of a New 1’1aenoZic Substance, Pg*atensol, C17H902(0H)3*The final sodium carbonate extract of the above-mentionedethereal liquid yielded, on acidification, a light brown powder.This was dissolved in hot alcohol, and the solution kept for sometime, when a small amount of the previously described pratol,C,,-J31204 (m. p. 253O), was deposited. The mother liquors fromhlie latter contained a quantity of a very soluble substance, whichcould not be obtained directly in a crystalline form.The alcoholwas therefore removed, and the residue acetylated, when a productwas obtained which, after a few crystallisations from alcohol,yielded a pure, colourless ucetyl derivative, melting at 189O :0.1071 gave 0.2569 CO, and 0.0445 H,O. C=65.4; H=4*6.O*1002 ,, 0.2410 CO, ,, 0*0406 H20. C=65*6; H=4.5.Tlie molecular weight of the acetyl derivative was determined0.3518, in 25-62 of benzene, gave At 0’094O.The number of acetyl groups in this compound was ascertainedby hydrolysing with aqueous sodium hydroxide, then adding aslight excess of sulphuric acid, and distilling the mixture in acurrent of steam:0,2057 gave acetic acid equivalent to 14.45 C.C.lV/10-Ba(OH)2.by the ebullioscopic method :M.W. =390.C23H1808 requires C = 65.4 ; H = 4.3 per cent. M.W. =423.CO*CH3 = 30.2.C,7H,0S(CO*CH3)3 requires CO-CH, = 30.6 per cent.From these results it may be concluded that the substance con-tained in red clover flowers, from which the above-described tri-acetyl derivative was prepared, possesses the formula C,,Hl20,.In order to obtain the parent compound, the acetyl derivative washeated for some time in alcoholic solution with dilute sulphuricacid, and the alcohol subsequently removed by distillation withsteam, when a colourless, crystalline substance separated. This waspurified by crystallisation from benzene, when it separated infeathery needles, melting a t 2 1 0 O .Although the amount of thissubstance w w insufficient for analysis, it was proved to represen238 POWER AND SALWAYthe original compound by the fact that when again acetylated, ityielded an m t y l derivative melting at 189O.No compound of the formula C,,H,,O,, possessing the sameproperties as that above described, appears to have hitherto beenrecorded. It is therefore proposed to designate the substame con-tained in red clover flowers as pratensol, with reference to itsphenolic character and the specific botanical name of the plant.Pratensol, C17Hg02(OH),, is very readily soluble in alcohol andacetic acid, but only sparingly soluble in water, chloroform, andbenzene. It dissolves in alkali carbonates, yielding yellow solutions,whilst its alcoholic solution gives with ferric chloride a greenish-black coloration.Triacetylp-at ensol, C17Hg0,(CO~CH3)3, crystallises from alcohol,in which it is only moderately soluble, in colourless, slender needles.It is readily soluble in glacial acetic acid, and moderately solublein benzene.When warmed with djlute aqueous sodium hydroxide,not only are the three acetyl groups removed, but the moleculealso undergoes further change, with the formation of a, compoundwhich crystallises from dilute alcohol in long, colourless needles,melting a t 183-184O. I f , on the other hand, the acetyl compoundis heated with aqueous sodium carbonate, the product is a yellow,crystalline substance, which melts and decomposes at 160O. It wasnot possible to ascertain the exact nature of these changes onaccount of the small amount of substance available, since 37.1kilograms of dried clover flowers yielded not more than 1.5 grams,or about 0.004 per cent., of pure triacetylpratensol.Isolation of a New Phenolic Substance, C15H703(0H)3.The sodium hydroxide extracts ( c ) of the previously mentionedethereal liquid, when acidified, yielded a light green powder, whichpossessed an odour resembling that of eugenol. An attempt toidentify the latter by distilling the product with steam wm, how-ever, unsuccessful.The light green powder was subsequentlydecolorised by treating its hot alcoholic solution with animalcharcoal, when, on cooling, colourless clusters of feathery needlesseparated. After several crystallisations, the substance wasobtained pure, and then melted at 225O:0.0931 gave 0.2133 GO2 and 0.0310 H,O.This compound yielded an acetyl derivative, which separatedfrom alcohol in thin, silky needles, melting a t 209O.The acetylcontent of this compound was ascertained by hydrolysing withdilute aqueous sodium hydroxide, then acidifying with sulphuricacid, and distilling the mixture in a current of steam:C = 62.5 ; H =3*7.C,,H1006 requires C = 62.9 ; H = 3.5 per centTHE CONSTlTUENT8 OF XED CLOVER FLOWERS. 2390.1750 gave acetic acid equivalent to 12.5 C.C. N/lO-Ba(OH),.CO*CH3 = 30.7.The molecular weight of the acetyl derivative was also ascertainedby the ebullioscopic method :0.1768, in 26.46 of benzene, gave A t 0-043O.C1,H,O,(CO*CH,), requires CO-CH,= 31.3 per cent.M.W. =412.The substance, Cl5Hl0O6, is soluble in alkali hydroxides, forminga colourless solution. I n alcoholic solution it gives with ferricchloride a dark green coloration. With concentrated sulphuricacid, only a faint yellow colour is produced.The yield of the substance C15H1006 was very small, not morethan 0.5 gram having been obtained from 37.1 kilograms of driedclover flowers.1solct.tion of a New Ghcoside, Trifolin, C22H220,,,H,0.The original aqueous liquid (A), which had been extracted withether, as already described, was kept foy a, considerable time, whenit was observed that a quantity (about 60 grams) of a light brownsolid had been deposited. This was collected, and purified byrepeated crystallisation from aqueous pyridine, when it separatedin minute, slender, pale yellow needles, melting and decomposinga t about 260O.The substance contained water of crystallisation,which was lost at 115O, but was again absorbed on exposing theanhydrous substance t o the air:M.W.=415.0.3051, on heating a t 115O, lost 0'0130 H20.0*1101* gave 0.2302 CO, and 0.0464 H20. C=57*0; H =4.7.0'1194* ,, 0.2481 (30, ,, 0*0506 H20. C=56*7; H=4*7.0*1010 * ,, 0.2115 CO, ,, 0.0440 H20. C=57.1; H=4*8.C2,H2,O,,,H20 requires H20 = 3.7 per cent.C22H2,011 requires C = 57.1 ; H = 4.8 per cent.H20=4'3.These results indicated the formula of the above-described sub-stance to be C22H22011, and it proved to be a glucoside. As it is anew compound, it is proposed to designate it trifolin, with referenceto the generic name of the plant from which it was obtained.Trif olin, C2,H2,OlI,H2O, is insoluble in chloroform, benzene, orcold water. It is not readily dissolved by alcohol, but is verysoluble in pyridine, and is best crystallised from a mixture of thelatter solvent with water.With aqueous sodium carbonate andthe alkali hydroxides, it gives intensely yellow solutions. It dis-solves in concentrated sulphuric acid, forming a t first a yellowsolution, but which rapidly develops a brilliant green fluorescence.I n alcoholic solution it gives with ferric chloride a dark browncoloration. No crystalline acetyl derivative of trifolin could beprepared.i+ Anhydrous substance240 POWER AND SALWBY:Hydimolysis of TiifoZin : Formution of u Yellow Colouriny Natter,T?if oli tin, C 16HTrifolin was dissolved in alcohol, it little dilute sulphuric acidadded, and the mixture heated on the water-bath for an hour.The alcohol was then removed by distillation in a current of steam,when a yellow, crystalline substance separated.This was collected,and washed with cold water, the filtrate having been set aside forthe subsequent examination of the sugar. Some difficulty wasexperienced in satisfactorily crystallising the yellow hydrolyticproduct, for it is extremely soluble in alcohol, and invariablyseparates from dilute alcohol at the ordinary temperature in anapparently amorphous condition. On the other hand, i f the alcoholis removed from the solution a t the boiling temperature, the sub-stance then separates from the hot liquid in thin, yellow needles.It is, however, best purified by crystallisation from moist nitro-benzene, when it separates in clusters of slender, yellow needles,which melt and decompose at about 275O.After drying a t 130°,it was analysed:0.0786 gave 0.1846 CO, and 0.0264 H,O. C = 64.1 ; H = 3.7.By heating the yellow hydrolytic product with acetic anhydridean acetyl derivative was obtained The latter, when crystallisedfrom alcohol, separated in colourless, silky needles, which, whenrapidly heated, melted a t about 116O, then resolidified, and finallymelted a t 182O. This behaviour was due to the presence of waterof crystallisation, the melting point of the anhydrous substancebeing 1 8 2 O :C,6H&6 requires C = 64.4 ; H = 3.4 per cent.0.5676, on heating at 105O, lost 0-0231 H,O.The anhydrous substance was analysed, and its molecular weight0.1078 gave 0.2426 CO, and 0.0401 H,O.0.4012, in 28.14 of phenol, gave ht-0*216°.C,,H6O6(C0*CH3), requires C= 61.8; H= 3.9 per cent.M.W. =466.In order to ascertain the number of acetyl groups in the abovccompound, it was dissolved in alcohol, dilute sulphuric acid added,and the mixture heated for several hours on the water-bath. Thealcohol was then completely removed by distillation in a current ofsteam, and, after cooling, the precipitated yellow substance wascollected and weighed, it having been found that the amountdissolved by the water was negligible:H,o=4'1.C,6Hti06(CO*CH3)4,H,0 requires H,O = 3.7 per cent.determined, with the following results :C = 61.4 ; H =4.1.M.W.=495.0.1056 ,, 0.2375 CO, ,, 0.0391 H20.C=61.3; H=4.1THE CONSTITUEN'I'S OF KED CLOVER FLOWERS. 2410*1811 gave 0.1160 of colouring matter, or 64.1 per cent.0.2922 ,, 0.1876 2, 9 , 64.2 ,,C1,HgO6(CO*CH3), requires C,,H,,O, = 63.9 per cent.The yellow colouring matter, as obtained from the above-described acetyl derivative, was crystallised from dilute alcohol,dried a t 120°, and again analysed:0.0855 gave 0.2000 CO, and 0.0276 H,O.C,,H,,06 requires C = 64.4 ; H = 3.4 per cent.It is evident from the above results that the yellow colouringmatter obtained by the hydrolysis of the glucoside trifolin possessesthe formula C,,H1,O,, and that it contains four hydroxyl groups.As no compound of this formula, having the same properties, hashitherto been recorded, it is proposed to designate it trifolitin.Trifolitin, C16H1006, is readily soluble in alcohol and glacialacetic acid, but only very sparingly soluble in benzene, chloroform,or ether.It dissolves in alkalis with an intense yellow colour, anddyes mordanted cotton wool a bright yellow. When dissolved inconcentrated sulphuric acid, it gives a yellow solution, followed bya brilliant green fluorescence. It is precipitated from its alcoholicsolution Sy basic lead acetate as an orange-yellow lead salt. I nalcoholic solution it gives with ferric chloride a dark greencoloration. Trifolitin is not altered when heated for several hourswith 30 per cent.aqueous potassium hydroxide, and it contains nomethoxyl group. It is evident from its empirical formula, CI6Hl0O6,that .it cannot belong to the flavone group of natural colouringmatters, and it differs, moreover, from the flavone compounds bythe fact that it forms no salt with sulphuric acid (Trans., 1896,80, 206), whilst it is only with difficulty that a potassium salt isdeposited from its concentrated alcoholic solution on the additionof potassium acetate. On the other hand, the composition andgeneral properties of trifolitin would agree with those of a tetra-hydroxy-derivative of phenylnaphthaquinone, but the amount ofmaterial availablc was not sufficient to enable us definiteIy toascertain the group of naturally occurring dyes to which it belongs.The aqueous liquid obtained in the course of hydrolysis of theglucoside trifolin was treated with barium carbonate for the removalof the sulphuric acid, and evaporated to the consistency of a syrup.This liquid readily reduced Fehling's solution, and yielded acrystalline osazone which, in the crude condition, melted anddecomposed a t 1 80°.When crystallised from aqueous pyridine, itseparated in needles, which melted and decomposed a t 194O:C=63*8; H=3*6.0.1258 gave 0.2902 CO, and 0.0738 H,O.The above result indicated that the sugar formed by theVOL. XCVJI. EtC = 62.9 ; H = 6.5.CI8H,,O3N, requires C = 63.2 ; H = 6.4 per cent242 POWER AND SALWAY:hydrolysis of trifolin has the formula C6H1205, and it appeared tobe rhamnose, the osazone of which, however, is stated t o melt a t180O.For the purpose of comparison, an osazone was preparedfrom a commercial specimen of rhamnose. This, when crystallisedfrom aqueous pyridine, was found to melt and decompose at.188O, and appeared to be identical with that above described.Trifolin is therefore converted by hydrolysis into trifolitin andrhamnose. A known weight of the glucoside was subsequentlyhydrolysed by dilufe sulphuric acid in aqueous alcohol, the alcoholthen removed by passing sfeam through the mixture, and, aftercooling, the precipitated trifolitin was collected, washed with coldwater, dried a t 120°, and weighed:0.2056 gave 0.1318 of hydrolytic product, or 64.1 per cent.0-2885 ,, 0-1859 9 , 7 9 64.4 ,,C22H220,, requires C16H1,06 = 64.5 per cent.The hydrolysis of trifolin therefore takes place according to thefollowing equation :C22H22011 = C16H1006 -k c6H1205*Trifolin.Trifolitin. Rhamnose.This equation would appear to indicate that the formation oftrif olitin and rhamnose from trifolin is unaccompanied by theabsorption of water, and would thus be analogous to that repre-senting the hydrolysis of quercitrin * :C21H22012 = C15H1007 + C6H1206*Quercitrin. Quercetin. Rhamnose.The explanation of this apparent anomaly may be, either thatthese glucosides tenaciously retain a molecule of water, or that therespective phenolic hydrolytic products suffer dehydration imme-diately after their formation.The dark-coloured aqueous liquid from which the above-describedglucoside, trifolin, had separated was next shaken repeatedly withhot amyl alcohol.The amyl alcohol extracts were united, washedwith a little water, and then concentrated under diminished pres-sure, when, on cooling, a yellow, gelatinous subdance was deposited.This was removed by filtration, and the filtrate again conceGtrated,when a further quantity of a yellow substance was obtained. Tothe final amyl-alcoholic filtrate, light petroleum was added, which* The hydrolysis of quercitrin is usually represented as fo!lows :This equation, Iiowever, is evidently illcorrect, inasmncli as it is now I ~ n o w nthat rhamiiose possesses the formula CGHl2O5, but crystallises with one nrolecnle ofwater (comlmre A.G. I’erkiu, J. Xoc. Chem. Ind., 1903, 22, 602, and Brauns, Amh.Phnrm., 1904, 24.2, 561).C2IHydOIZ i H,O=C,,H,o07 + CGHIjOGTHE CONSTITUENTS OF RED CLOVER FLOWERS. 243precipitated the remainder of the yellow substance as an amorphouspowder. The three fractions thus obtained were separatelyexamined.Fraction I.-This product consisted of a yellow, highly hygro-scopic substance. It was dissolved in hot alcohol, and the solutionkept for some time, when a gelatinous, brown mass separated. Thelatter was twice subjected to the preceding treatment with alcohol,when a crystalline substance was obtained, which was finally purifiedby crystallisation from dilute alcohol. It then formed colourless,glistening leaflets, melting at 214O. The amount of pure substanceisolated was 0.5 gram :0.0989 gme 0.2199 CO, and 0.0440 H,O.C = 60.6 ; H = 4.9.0'1154 ,, 0.2559 CO, ,, 0.0496 HZO. C=60.5; H=4.8.C,4H1206 requires C = 60.9 ; H = 4.3 per cent.This substance would thus appear to possess the formula Cl4HI2O6,and, as it differs in its properties from any substance of thisformula which has hitherto been recorded, it may be regarded asa new compound.The compound, C14H1206, is insoluble in water, chloroform, orbenzene, but readily soluble in hot alcohol. It is insoluble inaqueous sodium carbonate, but readily dissolves in alkali hydroxideswithout change of colour. It gives no coloration with concentratedsulphuric acid, and is not changed by heating with dilute sulphuricacid. Its alcoholic solution yields with ferric chloride a light greencoloration.Isolation of a Second Glucoside of TrifoEitin, isoTrifolin, C,,H,,O,,.Fraction ZZ.-This product, consisting of a light yellow powder,was fractionally crystallised from alcohol.The more solubleportion contained a substance which separated from very dilutealcohol in pale yellow needles, melting and decomposing at about250° with the formation of a deep red liquid. It was dried a t 115O :0.1100 gave 0.2282 CO, and 0.0451 H,O.The glucosidic character of this substance was proved by heatingwith dilute sulphuric acid in aqueous alcohol. On removing thealcohol, a yellow colouring matter separated from the hot solutionin thin needles, which melt.ed and decomposed at 275O. Thishydrolytic product yielded an acetyl derivative, which was foundto be identical with that prepared from the previously describedtrifolitin, C,,H,,O6.The aqueous liquid obtained in the course ofthe above hydrolysis readily reduced Fehling's solution, but theamount of glucoside available (0.5 gram) did not, permit of pre-paring an osazone from the sugar.C=56*6; H=4*6.C22H22011 requires C = 57.1 ; H = 4.8 per cent.R 244 POWER AND SALWAY :Although the above-described glucoside yields the same yellowhydrolytic product, and possesses apparently the same empiricalformula as the glucoside trifolin, the two compounds are evidentlynot identical. Thus trifolin (m. p. 260°) separates from dilutealcohol in small, globular aggregates of fine needles, whilst theglucoside, melting a t 250°, is much more soluble, and crystallises inclusters of comparatively large needles.It is therefore proposedto designate the second glucoside of trifolitin as isotrif olin.The general behaviour of isotrifolin, as might be expected, isvery similar to that of trifolin. It dissolves in alkalis with theformation of a deep yellow solution. With concentrated sulphuricacid, a yellow coloration is produced, together with a greenfluorescence, and in alcoholic solution i t gives with ferric chloride adeep brown colour.Isolation of u Gducoside of Quercetin.The sparingly soluble portion of the above-described fraction I1was finally purified by crystallisation from water, when it separatedin pale yellow, prismatic needles, melting and decomposing at 235O :0.1484, on heating at l l O o , lost 0.0161 H,O.0.1117, dried at l l O o , gave 0.2184 CO, and 0.0475 H,O.The substance dissolved readily in hot water, giving a yellowsolution, the colour of which was greatly intensified by the additionof alkalis. With concentrated sulphuric acid, a green fluorescencewas produced. I t s aqueous solution gave with ferric chloride adark green coloration.On heating the substance for a short time in aqueous solutionwith dilute sulphuric acid, a crystalline, yellow compound separated,and the filtrate from the latter readily reduced Fehling's solution.It was thus evident that the substance was a glucoside.The yellowhydrolytic product melted and decomposed at 305O, and, by meansof its acetyl derivative (m.p. 195O), was identified as quercetin.The amount of this glucoside obtained from 37.1 kilograms of driedred clover flowers was only about 0.5 gram, and it was thereforenot possible more definitely to daracterise it.Fraction ZZZ.-This was a yellow, amorphous solid, which wasfreely soluble in alcohol, giving a dark-coloured solution. Theattempts to obtain something crystallinc from this solution wereunsuccessful, and as a preliminary test indicated the presence ofglucosidic substances, the whole of the fraction, together with tlicmother liquors from fractions I a r d 11, was hydrolysed by heatingfor some time with dilute sulphuric acid in aqueous alcohol. Afterremoving the alcohol, a quantity of a dark-coloured solid separated,H20=10.8.C=53*3; H=4*7THE CONSTITUENTS OF RED CLOVER FLOWERS.245which was collected and washed with water. This product wasfound to be a highly coinplex mixture, which, however, by digestionwith ether, could be separated into readily and sparingly solubleportions. The ethereal solution of the readily soluble portion wasextracted with alkalis, when it yielded traces of salicylic acid, asmall amount of pratol, CI6Hl2O1, and a yellow coinpound meltinga t 305O, which, by means of its acetyl derivative (m. p. 194O), wasidentified as quercetin.Isolation of isoRhamnetin, Cl,H,,07.The portion of the above-mentioned hydrolytic product whichwas sparingly soluble in ether was treated with aqueous sodiumcarbonate, when it was found to be only partly soluble in thealkaline liquid.From the insoluble portion a, small amount ofthe previously described compound, CI4Hl2O6 (m. p. 214O), wasobtained, whilst the portion soluble in the alkali carbonate con-sisted chiefly of a yellow colouring matter, which was observed todiffer from any of the substances already isolated from the cloverflowers. This substance was sparingly soluble in acetic acid andalcohol; and was purified by conversion into its acetyl derivative,the latter being then fractionally crystallised from alcohol untilconstant in melting point. It then separated in thin, hair-likeneedles, melting at 198-200°. After being dried a t 120O:0.0927 gave 0.2017 CO, and 0.0354 H,O.CI6H8Oi(~*CH3), requires C = 59.5 ; H = 4.1 per cent,The number of acetyl groups in this derivative mas determinedby hydrolysis with dilute sulphuric acid, the hydrolytic productbeing then collected and weighed :C=59.3; H=4.2.0-1120 gave 0.0722 of colouring matter, or 64.5 per cent.0.1126 ,, 0.0734 ? 9 Y2 65.2 ,,Cl,H80,(CO~CH3), requires CI6H1,O7 = 65.3 per cent.The hydrolytic product, which melted and decomposed at about295O, was a deep yellow substance, only sparingly soluble in alcoholor acetic acid, and insoluble in chloroform or benzene.It dissolvedin alkalis, with the production of a deep yellow colour, and itssolution in sulphuric acid possessed a green fluorescence. It wasfound to contain one methoxyl group, as determined by Perkin’smodification of the Zeisel method :0.0941 of acetyl compound gave 0.0379 AgI.C,,H,O,(OCH,)(CO*CH,), requires Me0 = 6.4 per cent.From these analytical data it may be concluded that the yellowsubstance has the empirical formula Cl,H,,Oi, and that it containsfour hydroxyl groups and one methoxyl group.The compositionMe0 =5*3246 POWER AND SATJWAY:and characters of this substa.nce are thus in agreement with thoseof isorhamnetin, a colouring matter which was first isolated byA. G. Perkin (Trans., 1896, 69, 1658) from the petals of the yellowwallflower (Cheirnnthus Cheiri). The last-mentioned author had,however, recorded the melting point of tetra-acetylisorhamnetin asI n order to further confirm the identity of the substance con-tained in red clover flowers with isorhamnetin, the product resultingfrom the niethoxyl determination was examined.This separatedfrom the acid solution, on cooling, in yellow, silky needles, whichdid not melt, but partially volatilised, a t a temperature below3 0 0 O . It gave an acetyl derivative melting a t 190--192O, andpossessed the general characters of quercetin, thus affording con-clusive evidence that the respective substance in red clover flowerswas a monomethyl derivative of quercetin and identical with iso-rhamnetin.195-1 96'.The aqueous liquid which had been extracted with hot amylalcohol, as above described, was dark brown in colour, and gavean abundant yellow precipitate on the addition of basic leadacetate. This precipitate was examined in the usual way, but,with the exception of a small amount of yellow colouring matter,nothing definite could be isolated from it.The filtrate from thebasic lead acetate precipitate, after removal of the excess of lead,was of a bright yellow colour, and readily reduced Fehling'asolution. It, contained a large quantity of sugar, which yieldedd-phenylglucosazone, melting a t 2 0 5 O . The liquid was concentratedunder diminished pressure to t,he consistency of a, syrup, and keptfor a considerable time, but nothing crystalline separated from it.The Resins (B).The resinous material obtained from 1.5 kilograms of the originalalcoholic extract of red clover flowers was a dark green, soft solid,and amounted to 314 grams. It was dissolved in alcohol, thesolution mixed with purified sawdust, and the thoroughly driedmixture then successively extracted in a Soxhlet apparatus withvarious solvents, when the following amounts of extract., dried at100°, were obtained :Petroleum (h.p. 30-45") cstractod 190.0 grams = 60.5 1)er cent.Ether ) ) 19'0 ,, = 6'1 ,,Chloroform ,, 6.7 ,, = 2.1 ,,Alcohol ,, 43.0 ), = 13.7 ) )Ethyl Acetate ,, 10.3 ,, = 3.3 ,,- -Total 269'0 grams = 85.7 per centTHE CONSTITUENTS OF RED CLOVER FLOWERS. 247Petroleum Extract of the Resin.It washydrolysed by heating for a short time with an alcoholic solutionof potassium hydroxide, after which the alcohol was removed, wateradded, and the alkaline mixture repeatedly extracted with ether.The ethereal extracts were united, washed, dried, and the solventremoved, when a quantity (46.5 grams) of a brownish-yellow solidwas obtained.This extract consisted of a dark green, semi-solid fat.Isolation of .iVyricyl A Zcohol, C,,H,,*OH.The above-mentioned, brownish-yellow, unsaponifiable materialwas agitated with cold alcohol, and thus separated into a solubloand an insoluble portion. The latter consisted of a colourless, wax-like solid, melting somewhat indefinitely at 70-74O.It appearedto consist of a mixture of several substances, and was thereforesubjected t o prolonged fractional crystallisation from it mixtureof ethyl acetate and alcohol. I n this manner the least solubleconstituent was ultimately obtained pure, and then melted a t 85O.It crystallised from ethyl acetate in small, colourless needles, whilstfrom alcohol it separated in glistening leaflets :0.1250 gave 0.3776 CO, and 0.1597 HiO.The analysis and properties of this substance establish its identityC = 82.4 ; H = 14.2.C,,H,, requires c = 82.3 ; H = 14.2 per cent.as myricyl alcohol.Isolation of Heptacosane, C27H56, Hentriacontane, CZ1HG4, andSit 0s t ero I, C,,H,,O.The more soluble substances obtained in the course of the abovefractionation could not be further separated by crystallisation.Apreliminary examination having indicated that they consistedchiefly of hydrocarbons, with presumably a little myricyl alcohol,the mixture was heated at 130° for an hour with an equal weight ofphthalic anhxdride, in order to convert the myricyl alcohol into itsacid phthalic ester, and thus effect its removal.The product wassubsequently digested with a mixture of ether and chlorofoym, thesolution filtered, and shaken with aqueous sodium carbonate. Aninsoluble compound thus separated, which was collected, and foundto consist of myricyl sodium phthalate. The ether-chloroformsolution was then evaporated, and the residue heated with aqueoussodium hydroxide in order to remove unchanged phthalic anhydride.The portion insoluble in the alkali was afterwards dissolved inchloroform, the solution being washed, dried, and the solven248 POWER AND SALWAY:remcved. A wax-like solid was thus obtained, which was sub-jected to fractional crystallisation from ethyl acetate. The mostsparingly soluble fraction, which Separated in pearly leaflets, meltedat 65O, and this melting point was not changed by furthercrystallisation :0*1001 gave 0.3121 CO, and 0.1324 H,O.C,,H,, requires C = 85.3 ; H = 14.7 per cent.This substance was evidently hentriacontane (rn.p. 68O), thesomewhat low melting point being probably due to the presenceof a small amount of the following hydrocarbon.From the more readily soluble fractions, a substance was isolated,which separated in pearly leaflets, melting a t 58-59O:0,0944 gave 0.2943 CO, and 0.1229 H,O.C,,H,, requires C = 85.3 ; H = 14.7 per cent.This substance was thus identified as heptacosane.The portion of the unsaponifiable material which was soluble incold alcohol amounted t o 30.2 grams. It was twice distillea underdiminished pressure, and the following five fractions collected :200-210°; 210--220°; 220-290O; 290-300O; and above 300°/20mm.Fraction 200-210°/20 mm.-This was a pale yellow liquid, andcontained a considerable quantity of an unsaturated substance :0.1083 gaire 0.3256 CO, and 0.1289 H20. C =82.0; H = 13.2.0.1234 absorbed 0.1008 iodine.Fraction 210-220°/20 mm.-This was a brownish-yellow liquid,0.1064 gave 0.3141 CO, and 0.1288 H20.0.0990 absorbed 0.0832 iodine.Both the above fractions were evidently mixtures, and nothingof a definite nature could be obtained from them.Fractions 220--290° ; 290-300° ; and above 300°/20 mm.-Thesefractions solidified on cooling, and were found to consist chieflyof a phytosterol, with a small quantity of the previously mentionedhydrocarbons.After repeated crystallisation, a substance wasisolated which separated from a mixture of ethyl acetate andalcohol in colourless, glistening leaflets, melting a t 135-136O :C =85*0; H= 14-7.C=85.0; ;R=14*5.Iodine value = 81.7.and, like the preceding fraction, was unsaturated :C =80*5 ; H= 13.5.Iodine value = 84.0.0.3194, on heating at l l O o , lost 0.0160 H,O.0.0963, dried at l l O o , gave 0.2959 CO, and 0.1033 H,O.H,O=5.0.C,,H,,O,H,O requires H,O = 4.5 per cent.C = 83.8 ; H = 11.9.C27H46O requires C = 83.9 ; H = 11.9 per cent.0.2244 of anhydrous substance, made up to 25 C.C. with chloro-form, gave a, -Oo37/ in a 2 dcm. tube, whence [aID -34.4OTHE CONSTITUENTS OF RED CIAOVER FLOWERS.249The above-described substance was evidently a phytosterol, andit gave the colour reactions of this class of compounds. It yieldedan acetyl derivative, melting a t 12G0, and therefore i s identical mitJisitosterol (Monatsh., 1597, 18, 551).7solation of a New Dihydric Alcohol, Trifolinnol, C2,Hx02(OH)2.The aqueous alkaline solution of potassium salts resulting fromthe hydrolysis of the petroleum extract of the resin, having beenshaken with ether to remove unsaponifiable material, as alreadydescribed, was acidified with dilute sulphuric acid, when a dark-coloured semi-solid precipitate of fatty acids was produced. Onextracting the mixture with ether, however, it was observed that aportion of the precipitate was very sparingly soluble in thatliquid. The sparingly soluble material was therefore collected,when it formed a pale green solid, amounting to 3 grams.It waspurified by crystallisation from pyridine containing a little alcohol,from which it was deposited in minute, colourress needles, meltingand decomposing a t 295O. After being dried at l l O o :0.1049 gave 0.2746 CO, and 0.0982 H,O.This compound yielded an acetyl derivative, which separatedfrom alcohol in flat, colourless needles, melting at 165-166O. Ananalysis and the determination of its optical rotatory power gavethe following results :C = 71.4 ; H = 10.4.C21H3604 requires C = 71.6 ; H = 10.2 per cent.0.0968 gave 0.2442 GO2 and 0.0799 H20.0-2435, made up to 25 C.C. with chloroform, gave a, -Oo30/ in aFrom the above results it was evident that the compound meltingat 295O is a, dihydric alcohol, having the empirical formulaC2,H3604.As i t is a new substance, it is proposed to designateit trifolianol, with reference t o the generic name of the plant fromwhich it was obtained.5"rifoZianoZ, C,,H,,O,( OH),, is practically insoluble in water, andvery sparingly soluble in alcohol, ether, and chloroform, but readilysoluble in dilute pyridine, from which it crystallises in minuteneedles. When dissolved in chloroform with a little aceticanhydride, and a drop of concentrated salphuric acid subsequentlyadded, it gives a transient pink coloration, changing to blue, andfinally to gregn. The properties of trifolianol are similar to thoseof two other dihydric alcohols which have recently been isolated inthese laboratories, namely, ipuranol, C23H,80,(OH)z ( A nzer.,7.PJmrnz., 1908, 80, 264, 576; Trans., 19OS, 93, 907; 1909, 95, 249),C = 68.8 ; H = 9.2.C,,H,,O,(CO*CH,), requires C = 68.8 ; H = 9.2 per cent.2-dcm. tube, whence [a], -25'7O250 POWER AND SALWAY:and citrullol, C,,H3,02(OH), (this vol., p. 102). These alcoholstherefore appear to be members of a homologous series, which isrepresented by the general formula, C,,H2, ,@.,.K,ramin,ation of the Fatty A cids.The ethereal solution, from which the above-described trifolianolhad been removed by filtration, was washed, dried, and the solventremoved. The mixture of fatty acids thus obtained contained aconsiderable quantity of resinous material, which was insoluble inlight petroleum, and was therefore removed by digesting themixture with the latter solvent, and filtering.This resin was Bdark green solid, amounting to 10 grams, but no definite crystallinecompound could be isolated from it. The petroleum solution, onevaporation, yielded a dark-coloured mixture of fatty acids, whichwas distilled under diminished pressure. The pale yellow distillate,amounting to 60 grams, became partly solid on cooling. It wasdissolved in alcohol, when a substance separated which, after severalcrystallisations from alcohol, melted at GO-GlO :0.1182 gave 0.3236 CO, and 0.1341 H20. C=74.7; H=12.6.0.1280 required for neutralisation 5-02 C.C. N / 10-KOH.Neutralisation value= 220.C,,H3,02 requires C = 75.0 ; H = 12.5 per cent.Neutralisation value = 219.1.This substance was thus identified as palmitic acid.I n order to ascertain the nature of the remaining constituents ofthe mixed fatty acids, the mother liquors from the abovecrystallisations were united, and, while hot, fractionally precipitatedwith a concentrated aqueous solution of barium acetate.The finalprecipita.te was oily, and, as it consisted for the most part of thebarium salts of unsaturated acids, it was separately examined asdescribed below. From the other fractions of barium salt the fattyacids were regenerated, and, after one crystallisation from aceticacid, were titrated with a decinormal solution of potassiumhydroxide. The neutralisation values of the several fractions were199.0; 210.0; 213.0; 213.5; 214.8; and 216.9, whilst their meltingpoints were 53-55O; 53-56O; 53-56O; 53-56O; 54-56O; and5G-59O respectively.These results indicated that the satdratedacids consisted chiefly of a mixture of palmitic and stearic acids,the respective neutralisation values of which are 219.1 and 197.7.The acids obtained from the above-mentioned oily barium saltwere converted into their lead salts, and the latter treated withether. From the portion of lead salt which was undissolved bythe ether, a further small amount of palmitic acid was obtained.The acids regenerated from the soluble lead salt formed a palTHE COKSTITUENTS OF RED CLOVER FLOWERS. 251yellow oil, which was distilled under diminished pressure. Ananalysis and the determination of its constants gave the followingresults :0.1428 gave 0.4031 CO, and 0.1437 H,O.d 2Oo/2O0 = 0.9148. Neutralisation value 200.9.Iodine value 184.7.CI8H3,O2 requires C = 76-6 ; H = 12.1 per cent. NeutralisationCIBH32O2 requires C = 77.1 ; H = 11.4 per cent. NeutralisationC,,H,,O, requires C = 77.7 ; H = 10.8 per cent.. NeutralisationI n order to ascertain more definitely the nature of the unsaturatedacids, 10 grams of the mixture were dissolved in alkali, andoxidised with a 1.5 per cent. solution of potassium permanganate.The chief product of the oxidation, when crystallised from water,separated in thin needles, melting at 156-157O:C = 77.0; H = 11.2.value = 198.9.value= 200.4. Iodine value = 181.4.value= 201.8.Iodine value = 90.1.Iodine value = 274.0.0.0630 ga8ve 0.1426 CO, and 0.0594 H,O.This substance was evidently tetrahydroxystearic acid (sativicacid).I n addition to the latter acid, very small quantities ofdihydroxystearic acid (m. p. 130°), linusic acid (m. p. 200-205°),and isolinusic acid (m. p. 173-174O) were obtained.The above results would therefore indicate that the unsaturatedacids consisted chiefly of linolic acid, with smaller amounts of oleic,linolenic, and isolinolenic acids.C = 61.7 ; H = 10.5.C,,H3,0, requires C = 62.1 ; H = 10.3 per cent.Ethereal Extract of the Resin.This extract was a black, brittle solid, and amounted to 19 grams.It was digested with 500 C.C. of ether, when the greater portionpassed into solution, while a small amount of a light green powderremained undissolved. The latter was collected, and boiled withalcohol, in which it was very sparingly soluble, in order to removethe green colouring matter.It was finally crystallised from amixture of pyridine and alcohol, when it separated in colourlessneedles, melting and decomposing at about 290O. It yielded anacetyl derivative, melting a t 166O, which was analysed :0.0854 gave 0-2142 CO, and 0.0690 H,O.C2,H3,O,(CO*CH3), requires C = 68.8 ; H = 9.2 per cent.The above-described substance, when dissolved in chloroform witha little acetic anhydride, and a drop of concentrated sulphuric acidsubsequently added, gave a pink coloration, rapidly changing toblue, and finally to green. It was evidently identical with theC = 68.4 ; H = 9.0252 POWER AND SALWAY:dihydric alcohol designated as trif olimol, which had previouslybeen isolated from the petroleum sxtract of the resin.The above-mentioned ethereal solution of the ethereal extract,was shaken successively with aqueous ammonium carbonate, sodiumcarbonate, and sodium hydroxide, and the various extracts thusobtained were separately examined.The ammonium carbonateremoved only traces of a green resin. The first sodium carbonateextract yielded, on acidification, a dark green resin, which consistedchiefly of chlorophyll, whilst the subsequent sodium carbonateextract, when acidified, gave a nearly colourless precipitate. Thiswas collected and dissolved in alcohol, when a small amount of asubstance separated in colourless, talon-like needles, melting a t253O.This substance was found to be identical with the compounddesignated as pratol, CI6Hl2O4, which had previously been isolatedfrom the portion of the alcoholic extract of clover flowers whichwas soluble in water. The mother liquors from the crystallisationof the pratol deposited a little of a crystalline substance, meltingsomewhat indefinitely between 1 7 8 O and 183O, but the amount wastoo small to permit of its further examination. The final extractionof the ethereal liquid with sodium hydroxide yielded nothing butamorphous products, and on subsequently evaporating the etheronly traces of a soft resin remained.Chloroforna and EthLyl Acetate Extracts of the Resin.These extracts were black, brittle solids.They were small inamount, and nothing crystalline could be isolated from them,Alcohol Extruct of t h e Resin.This was a dark brown, brittle solid, amounting to 43 grams.I n order to ascertain whether it contained anything glucosidic, it,mas heated for some time with a 5 per cent. solution of sulphuricacid in aqueous alcohol, and the mixture subsequently distilledwith steam, but no volatile oil or acid passed over. The distillationflask then contained, besides the aqueous liquid, a quantity of ablack resin, which was collected and examined, but nothing definitecould be isolated from it. The aqueous liquid, on the other hand,when extracted with ether, yielded a small quantity of a viscid,vellow oil, which gradually deposited a solid substance.The latterwas crystallised from alcohol, when it separated in colourlessneedles, melting at 250°, and gave an acetyl derivative meltingat 166O. It proved to be identical with the previously describedsubstance, C,,H,,04, designated as pratol. The aqueous liquidTHE CONSTITUENTS OF RED CLOVER FLOWERS. 253which had been extracted with ether as above described, was freedfrom sulphuric acid by means of baryta, and concentrated to asmall volume. It then readily reduced Fehling's solution, andyielded a small quantity of an osazone, melting a t 203'.These facts indicated the presence of a glucoside in the alcoholicextract of the resin, and it is probable that the small amount ofpratol, ClGHl2O4, obtained from the latter represents one of itshydrolytic products.Sum?nary.The detailed results of the present investigation of red cloverflowers (Trifolium pratense, Linn6) may be summarised as follows :A quantity of the alcoholic extract, representing 37.1 kilogramsof the dried flowers, when distilled with steam, yielded 10.5 grams,or 0.028 per cent., of essential oil. This essential oil, whichcontained furfuraldehyde, possessed the following constants :d 2O0/2O0=0-9476; uD +4OO/ in a 1-dcm. tube.The portion of the alcoholic extract which was soluble in watercontained a large amount of sugar, which yielded d-phenyl-glucosazone (m. p. 205O), but from the aqueous liquid the followingdefinite substances were isolated : salicylic and p-coumaric acids ;isorhamnetin, CIGH1207 (m. p. 2 9 5 O ) , which wits probably presentin the form of a glucoside; a number of new phenolic substances:pratol, Cl,H802('OH)(OCH3), m. p. 253O, which apparent,ly is ahydroxymethoxyflavone, and yields an acetyl derivative melting a t166O; pratelzsol, C17H,02(OH),, m. p. 210°, which yields a triucetylderivative melting a t 189'; a yellow compound, C16H,,0, (m. p.280°), which yields a tetra-acetyl derivative melting at 145-147O;a substance, C15H703(OH)3, m. p. 225O, which yields a triacetglderivative melting at 209O; a substance, C14H,,0, (m. p. 214O);also the following new glucosides : trifolin, C2nH220,,,H20 (m. p.260°), which yields on hydrolysis a yellow colouring matter,trifolitin, C,,H,,O6 (m. p. 275O), and rhamnose, C6H1,0,; iso-trifolin, C22H220,1 (m. p. 250°), and a glucoside of quercetin, melt-ing at 235O.The portion of the alcoholic extract which was insoluble inwater consisted chiefly of resinous material, the amount of whichwas equivalent to 5.6, per cent. of the weight of dried clover flowers.From the resin the following compounds were obtained : myricylalcohol, C,,Ho,*OH ; heptacosane, C2,HGG, and hentriacontane,CY1HG.I; sitosterol, C27H460 (m. p. 135-136O; [ U ] ~ -3414"); a newdihydric alcohol, trifolianol, C,,H3,0,(0H), (m. p. 295O), whichappears to be a homologue of ipuranol, C2,H,802(0H),, and yieldsa diacetgl derivative melting at 165-166O and having [aID - 25'7O ;a mixture of fatty acids, consisting chiefly of palmitic, stearic, an254 FORSTER AND JUDD: THE TRIAZO-GROUP. PART X l I ,linolic acids, with small amounts of oleic, linolenic, and isolinolenicacids. A small amount of pratol, C16H1204, was likewise obtained,the latter having evidently been present in the resin in the formof a glucoside.THE WELLCOME CHEMICAL RESEARCH LABOILATO~LIES,LONDON, E.C