THE CONSTITUENTS OF SOLANUM ANGUSTIFOLIUM. 55 9LVII.-The Constituents o f Solanurn Angustgolium :Isolation of u iVew Gluco-alkuloid, Solungustine.By FRANK TUTIN and HUBERT WILLIAM BENTLEY CLEWER.IN several countries in South America, namely, Peru, Bolivia,Paraguay, and the southern portion of the province of BuenosAyres, in the Argentine Republic, a solanaceous plant occurs whichis known as “Duraznillo Blanco.” This plant, which has beenidentified as Solanum angustifolium, Ruiz et Pavon, was brought tothe notice of Dr. Power by Dr. E. H. Colbeck, as being a drugworthy of chemical investigation. It is employed in South Americaas a febrifuge, chiefly in the treatment of entjeric fever. I n Peru itis also used in cases of malaria, but with caution, on account of itsreputed poisonous properties.At the suggestion of Dr.Power, the present authors have there-fore conducted a chemical investigation of the drug in question,which has led to the isolation of a number of compounds, includinga new and interesting gluco-alkaloid.Several species of the extensive genus Solanurn have hithertobeen reported to contain bases which, in addition to their alkaloidalnature, were also glucosides. The name solanine has in all casesbeen applied to such bases, but a perusal of the literature of thissubject reveals a state of great confusion, and makes it appeardoubtful whether any pure gluco-alkaloid has heretofore beenisolated. Thus, the solanine from Solanum Dulcamara has beenstated to possess the formula C42H87015N o r C)5”?Kg2OI8N, and toyield, on hydrolysis, the base, solanidine, CzGH4,O2N. Solanine f rointhe shoots of the potato (8.tuberosum) has been stated t o yieldsolanidine having the formula C40H6102N.More recently, Odd0 and Colombano (Gazzetta, 1905, 35, i, 27)prepared solanine from S. sodomaum, and assigned t o it theformula (C23H3g0,N),,H,0. They stated that it yielded, on hydro-lysis, solanidine, C19H2gON, but the sugar that was also found wasnot identified. A t a later date, the same authors ( A t t i R. Accad.Lincei, 1906, [v], 15, ii, 312) modified their formula for solanineto (C,7H470gN)2,H20. Solanine from the seeds of 8. tuberosum wasthen investigated by Colombano (ibid., 1907, [v], 16, ii, 683), whostated that it differed from the base obtained from S.sodomceum,and had the formula C,,H,,O,,N. The most recent work on thesubject is by Odd0 and Cesaris (Gazzetta, 1911, 41, i, 490), whopropose to designate the bases obtained from the lasbmentioned twospecies of Solanurn as solanine-t and solanine-s respectively. Th560 TUTIN AND CLEN‘EK.latter compound they state to have the formula (C,7~4609N)2,H20,and represent its hydrolysis by the following equ B t’ ion:(C2,H4609N)2,H20 + H 2 0 +H, =Solanine-s.2Cl8H,,ON + C6HlZO, + C6H1,0, + C6H1205.Solanidine. Galactose. ? Dextrose. Illhamuosc.The base described in the present communication, which has beendesignated solangustine, is a t once differentiated from anypreviously described solanines by the insolubility of its salts.Itis considered that it has been definitely established that its forniulais C3,H,,07N, and that, on hydrolysis, i t yields solangustidim,C,,H,,O,N, together with one molecule of dextrose.A summary of the results of the general iiivestigation of theplant will be found a t the end of this paper.EXPERIMENTAL.The material employed in this investigation was obtained froinLima, Peru, and consisted of the leaves, twigs, and flowers of theplant which is there known as “Duraznillo Blanco.” It wasbotanically identified by Mr. E. M. Holmes, F.L.S., as Solarnuinangustifolium, Ruiz et Pavon.A small portion (10 grams) of the dried and ground materialwas digested with Prollius’ fluid, and the resulting extract testedfor an alkaloid. Copious precipitates were then obtained with allthe usual reagents, thus indicating the presence of a considerableproportion of alkaloidal material.Another portion (25 grams) of the dried and ground materialwas extracted successively in a Soxlilet apparatus with varioussolvents, when the following amounts of extract,, dried a t loo”,were obtained :Petroleum (b.p. 35-50’) extractxd 0’60 gram = 2.40 per cent.Chloroform ), 0.36 ,, = 1’44 ,,Ethyl acetate ,, 0.41 ,, = 1’64 ,,Alcohol ), 5.96 ,, = 23.84 ,,Ether ), 0.44 ) ) = 1.76 ) )-- -Total = 7-77 grains= 31 08 per cent.Por the purpose of a complete examination, 30.92 kilograms ofthe dried material were completely extracted with hot alcohol,when, after the removal of the greater part of the solvent,12.73 kilograms of a viscid, dark green extract were obtained.A quantity (3 kilograms) of the above-mentioned extract wasmixed with water, and steam passed through the mixture for severalhours, when only a slight trace of essential oil passed over.Therethen remained in the distillation flask a dark brown, aqueouTHE CONSTITUENTS OF SOLANUM ANGUYTIFOLIUM. 561liquid (A), and a quantity of a soft, dark green resin (B). Thelatter was separated, and repeatedly washed with boiling water,the combined washings being concentrated, and added to the mainbulk of the aqueous liquid.Examination of the Aqueous Liquid (A).It having been ascertained that the extraction of the aqueousliquid (A) with ether could only be slowly and with difficultyeffected, recourse was had to the use of chloroform.After tenextractions ,with the latter solvent, practically nothing more wasremoved, and the extracts, all of which were green, were washedand evaporated. The dark green, resinous residue was dissolvedin a small amount of alcohol, and then largely diluted with ether,when a small amount of dark green, resinous material was pre-cipitated and removed. The latter apparently contained a smallamount of amorphous, alkaloidal material, but nothing definitecould be obtained from it. The ethereal filtrate was shaken witheight successive portions of 1 per cent. hydrochloric acid, when theresulting extracts each yielded reactions for alkaloid. The extrac-tion was then continued with 5 per cent. hydrochloric acid untilthe basic substance ceased to be removed.The extrach so obtainedwere separately examined, but they yielded only brown, amorphousproducts, possessing a strong, basic odour. The amount of theseamorphous, alkaloidal products was not large, and no crystallinesalt could be obtained from them.The first liquid, which had been obtained by extraction with1 per cent. acid, as above described, after being made alkaline anddeprived of base, was re-acidified and extracted many times withether. A small amount of a product was thus obtained whichdeposited some crystals and gave the colour reactions of gallic acid,but the amount was too small f o r further investigation.The ethereal solution, which had been extracted with hydro-chloric acid, as above described, was then shaken successively withaqueous ammonium carbonate, sodium carbonate, and potassiumhydroxide. The first- and lasbmentioned alkalis removed onlysmall amounts of resin, but the sodium carbonate extract yieldeda little slightly impure quercetin, very much larger amounts ofwhich were subsequently shown to be present in the form of thegluco-rhamnoside, rutin.The neutral portion of the ethereal extract of the aqueous liquidconsisted of about 2 grams of dark green, fatty material562 TUTIN AND CIAEWER:The aqueous liquid (A), which had been extracted with chloro-form, as above described, was kept for a few days, when a quantity(6 grams) of a substance was deposited in fan-shaped tufts of lightyellow, microscopic crystals.This substance was very sparinglysoluble in alcohol, and practically insoluble in the other usualorganic solvents, with the exception of pyridine, in which it dis-solved readily. It was most conveniently purified by crystallisa-tion from water, in which it is sparingly soluble. On heating itsintered a t 170°, and melted indefinitely between 185O and 2 2 0 O .(Found, C = 48.9 ; H = 5.7. C,,H3,0,6,3H20 requires C= 48.8 ;I - = 5 * 4 per cent.)This substance was identified as the gluco-rhamnoside, rutin ;this has been shown t o crystallise with 3 molecules of water, whichare somewhat difficultly eliminated on heating (compare Clarke,T., 1910, 97, 1833). It has been shown by A. G. Perkin (ibid.,p. 1776) that the glucosides previously known as osyritrin,myrticolorin, and violaquercetrin, respectively, are also identicalwith rutin.A quantity of the rutin was hydrolysed by heating with dilute,aqueous sulphuric acid, and the resulting liquid extracted withether.Quercetin was then obtained in small, yellow needles, melt-ing a t 314O. (Found, C=59*4; H=3*6. Calc., C=59.6; H=3*3per cent.) It yielded penta-acetylquercetin, which formed soft,colourless needles, melting a t 194O. The acid, aqueous liquid fromwhich the quercetin had been removed was deprived of sulphuricacid and concentrated. On heating with phenylhydrazine acetate,i t readily yielded an osazone, which was collected in two successivefractions. The first fraction, after recrystallisation from dilutepyridine, melted a t 216O, and proved to be glucosazone.The secondfraction was somewhat readily soluble in alcohol, and, after re-crystallisation from this solvent, melted at 18Z0, and was foundto consist of rhamnosazone.Further amounts of rutin were subsequently isolated, as describedbelow, and the total amount obtained was about 55 grams, beingthus equivalent to about 0.75 per cent. of the weight of the plantemployed.The aqueous liquid (A) from which the rutin had been separatedwas extracted fifteen times with amyl alcohol, the resulting liquidsbeing washed and concentrated to a small bulk under diminishedpressure. The liquid obtained by the first extraction yielded onlya quantity of a dark-coloured, viscid product, readily soluble iTHE CONSTITUENTS OF SOLANUM ANGUSTIFOLIUM.563dry amyl alcohol. On concentrating the subsequent extracts, how-ever, a yellow, sparingly soluble solid separated. The latter wascollected, and the material in the filtrate added to the above-mentioned viscid product. I n this way two products wereobtained: (a) the material sparingly soluble in dry amyl alcohol,and ( b ) that readily dissolved by the latter.Examination of the Product (a).The product ( a ) was mixed with water and deprived of amylalcohol by means of steam, the resulting dark brown, aqueoussolution being freed from traces of green material by extractionwith ether. No crystalline substance, however, could be directlyobtained from this solution of the product (a), and it was thereforesought t o obtain some definite compound from it by hydrolysis.Such an amount of sulphuric acid was therefore added to theaqueous solution as to represent about 7.5 per cent.of the totalmixture, whereupon an insoluble product separated in semi-gelatinous granules. The mixture was then warmed on a water-bath for fifteen minutes, after which the dark-coloured solid wascollected on a filter. The filtrate was boiled for one hour in ordert o complete the hydrolysis of any glucoside i t might contain. Onshaking the cooled mixture with ether, and fractionally extractingthe resulting ethereal liquid with various alkalis, a quantity ofquercetin (m. p. 310°), together with some amorphous products,was obtained. It is evident, therefore, that either some rutin, oranother glucoside of quercetin, had been extracted by means of theainyl alcohol, although the latter solvent, in a pure state, does riotappear to dissolve rutin.1solatio)t of 3 : 4-UiAy~l~oryciri~iccniic A cicl.The acid, aqueous liquid which had been extracted with ether, asabove described, was made strongly alkaline with potassiumhydroxide, and boiled for a short time.The liquid was thenquickly acidified and cooled, after which it was extracted withether. On shaking the resulting ethereal liquid with aqueousammonium carbonate, and subsequently acidifying the alkalineliquid, a quantity (4 grams) of a crystalline acid was obtained.The latter separated from water in pale brown prisms, melting a t216O, and was identified as 3 : 4-dihydroxycinnamic acid.(Found,C = 602 ; H =4.6. It yielded3 : 4-dimethoxycinnamic acid, melting a t 180-181°.Calc., C = 60.0; H =4*4 per cent.)VOL. cv. P 564 TUTIN AND CLEWER:Isolutioib of a New Gluco-alkaloid, Solaiigustitie, C3.H,,07N,H,0.The previously mentioned dark-coloured solid which hadseparated on the addition of sulphuric acid to the aqueous solutionof the product (a), as above described, was well washed with boil-ing alcohol, which removed most of the colour. It was theiidigested for some time with slightly diluted acetic acid, when, with-out dissolving to an appreciable extent, it eventually assumed acrystalline form. This substance proved to be the sulphate of analkaloid. I n working up the remainder of the original extract ofthe plant for the sole purpose of isolating a further quantity ofthis sulphate, it was not found necessary to follow the above-described procedure in full.The extract was mixed with waterand distilled in a current of steam for the removal of the alcohol.The resin was then separated from the aqueous liquid, and thelatter extracted many times with amyl alcohol, the extracts beingsubsequently washed with water and concentrated t o a small bulk.The resulting product, part of which consisted of solid materialwhich separated during the concentration, was then mixed withwater, and deprived of amyl alcohol by means of steam. Thedark-coloured, aqueous liquid so obtained was afterwards extractedrepeatedly with ether until the greater part of the material solublein this solvent was removed. It was then treated with sulphuricacid (about 5 per cent.of the weight of the liquid), the mixturewarmed gently for about fifteen minutes, cooled, and the precipi-tated sulphate of the alkaloid collected. The latter was finallypurified in the manner already indicated.I n order to obtain the base from its sulphate, it was necessaryto resort to the employment of warm amyl alcohol, since the alka-loid is insoluble, or nearly so, in all other usual solvents, and thesalt insoluble, or practically so, in everything. The sulpliate wastherefore mix& with warm, aqueous sodium carbonate, and themixture vigorously shaken with successive portions of warm arnylalcohol. The amyl alcohol extracts were then washed, and con-centrated to a small bulk under diminished pressure, when, on cool-ing, the alkaloid separated in the form of hard, pale yellow crusts,which, under the microscope, were seen to consist of aggregates ofsmall crystals.The base so obtained was found to be a newgluco-alkaloid, and has been designated solangustine, with refer-ence to its botanical source. The amount of it isolated corresponded with 0.062 per cent. of the weight of the air-dried drugemployed. Solangustine was recrystallised by dissolving it in alarge volume of hot aniyl alcohol, and then concentrating andcooling the solution. On heating, i t darkens slightly a t about 225OTHE CONSTITUENTS OF SOLASUM ANGUSTIFOLIUM. 565and melts and decoinposes a t 235O. It contains solvent of crystal-lisation, and, when dehydrated, rapidly absorbs water from theatmosphere :0.7457,* on heating a t 130°, lost 0.0233 H,O, after which, onH,O = 3.1.0.0991 * gave 0.2421 CO, and 0.0870 H,O.C=66.6; H=9.7.0.0980 * ,, 0'2386 CO, ,, 0.0832 H20. C=66*4; H=9*4.0.3019 * ,, 6.8 C.C. N, (moist) a t 1 8 O and 754 mm. N=2*6.C33H5307N,H20 requires C = 66.8 ; H = 9.3 ; N = 2-4 ;H20 = 3.0 per cent.It thus appears that solangustine possesses the f orrnula C,,H,,07N,and crystallises with 1 molecule of water, and this conclusion wasborne out by the analysis of its derivatives, described below. Theonly solvent in which solangustine will dissolve a t all readily ispyridine, but it cannot be crystallised from this liquid. Solangus-tine contains no methoxyl group, and its acetyl derivative wasfound to be uncrystallisable.Solungustine Sulphute, (C~H,307N),,H2S0,,3H20.-This salt wasprepared in a state of purity by shaking a solution of the respectivebase in amyl alcohol with dilute, aqueous sulphuric acid.A pre-cipitate then separated, which consisted of small, colourless, acicularcrystals, which did not melt or decompose a t 325O. Solangustinesulphate, like the corresponding base, contains water of crystallisa-tion, and, when dehydrated, it is extremely hygroscopic :exposure to the air, it reabsorbed 0.0231 H20.0.2144, on heating a t 140°, lost 0.0089 q0.0.1583 gave 0.0277 BaS04.H,O=4'1."SO, = 7.2.(CBH5307N)2,H2S04,3H20 requires H20 = 4.1 ; "SO, = 7.4 per cent.Solangustine sulphate appears to be insoluble in all solvents withthe exception of acetic acid, in which it dissolves very sparingly onboiling.No water-soluble salt of solangustine could be obtained, and theabove-described sulphate was the only salt isolated in a crystallinecondition.The hydrochloride and nitrate were amorphous, in-soluble products.Hydrolysis of Solnizgustirze.Formation of Solangustidine, C,,H4,O2N, and Dextrose.A quantity (5 grams) of solangustine was dissolved in 500 c.cof warm amyl alcohol, and 50 C.C. of 15 per cent. hydrochloric acicwere added, together with sufficient alcohol to render the mixturchomogeneous. The liquid was then boiled for six hours, after which* Heated at 130" t o expel amyl alcohol a i d then exposed to the air unticon stall t.P P 566 TUTIN AND CLEWER:i t was cooled, shaken with - a inoderate voluuie of water, and theaqueous layer separated. The latter was found to contain sugar,but no salt of an alkaloid.As, however, the amount of sugarformed was so small as to indicate that hydrolysis had not, beencomplete, fresh quantities of hydrochloric acid and alcohol wereadded to the amyl alcohol liquid, and the mixture again boiled forsix hours. It was found necessary to repeat this treatment fourtimes before sugar ceased t o be formed.The aqueous liquids containing sugar were then made exactlyneutral by the cautious addition of potassium hydroxide, afterwhich they were evaporated t o a small bulk under diminishedpressure. The greater part of the inorganic salt was then removedby precipitation with absolute alcohol, after which the f i h a t e wasdeprived of alcohol, and heated for two hours with aqueous phenyl-Iiydrazine acetate.The osazone which formed was collected andcarefully examined for the presence of rhamnosazone by Perkin’smethod (T., 1910, 97, 1777), when it was found to consist solely ofdextrosazone, melting a t 215O. It is thus evident that solangustinebelongs to the little-known group of gluco-alkaloids.Solnngztstidiiz e Hydrochloride, CT27H,30,N,HCl.-Tlie amyl alcoholsolution from wliicli the sugar had been removed by shaking withwater, as above described, was mixed with water and deprived ofamyl alcohol by means of steam. There then remained in the flaskan aqueous liquid, together with a quantity of a whib solid insuspension.The latter was collected, arid crystallised from absolutealcohol, to which a little alcoholic hydrogen chloride had beenadded, when it formed lustrousplates, which did not melt a t 3 2 5 O .It was most readily obtained crysballine by allowing its solution inboiling alcohol t o evaporate :0.0971 gave 0.2567 CO, and 0.0874 H,O.0.2047 ,, 0.0658 AgC1.t Cl=8*0.0’1913 ,, 0.0609 AgCl. C1=7.9.C,7H,,0,N,HCl requires C = 72.1 ; H = 9.8 ; C1= 7.9 per cent.This substance is thus seen to be the hydrochloride of a basewhich is produced, together with 1 molecule of dextrose, by thehydrolysis of solangustine. It is proposed to designate the hydro-lytic base solungustidine, and its formation may be represented byC=72.1; H=10-0:*Solungustidin e ?hydrochloride is sparingly soluble in amyl alcoholand hot ethyl alcohol, but is insoluble, or practically so, in the otherusual solvents.When i t is dis- It is quite insoluble in water.* Otlicr aiialyses gave C=71.7, 71.8; H = 9 * 9 , 10.0.t By fusion method. $ By Carius’ methodTHE CONSTITUENTS OF SOLBNUM ANGUSTIFOLIUM. 567solved in concentrated sulphuric acid, and the solution kept forsome time, a reddish-yellow, slightly fluorescent liquid is obtained.I n order to isolate solangustidine from its hydrochloride, aquantity of the latter was dissolved in alcohol, and the solutionmade alkaline by the addition of an alcoholic solution of sodiumethoxide. Water was then added, and the precipitated basecollected.As thus obtained, solangustidine was amorphous, andhad no definite melting point. When dry it formed a horn-likemass. It separated from dilute alcohol in amorphous granules, andall attempts to obtain it crystalline resulted in failure. It hadevidently suffered no change by tlie treatment with alkali, since itreadily regenerated tlie crystalline hydrochloride.Solangustidine Hydrobromide, Cz7H,,02N,HBr.-A quantity ofsolangustidine was dissolved in alcohol, and a solution of hydrogenbromide in glacial acetic acid added. The mixture was then con-centrated, when a colourless, crystalline solid separated. The latterwas dissolved in alcohol containing a little hydrogen bromide, andthe solution concentrated, when colourless, lustrous plates, whichmelted and decomposed a t 320°, separated from the boiling liquid.The hydrobromide is somewhat more soluble in alcohol than thecorresponding hydrochloride, but, like the latter, is quite insolublei n water :0.0892 gave 0.2132 CO, and 0.0749 H20.Cz7H,,0,N,HBr requires C = 65.6 ; H = 9.0 per cent.Solangustidin e Nitrate, C,7H,,0,N,HN03.-An alcoholic solutionof solangustidine was acidified with dilute nitric acid, after whichthe liquid was diluted with water until a turbidity was produced.The mixture was then warmed, and allowed t o cool slowly, whencolourless leaflets separated, which become brown a t 260° and meltand decompose a t 290O.The nitrate is practically insoluble inwater, but fairly readily soluble in hot, dilute alcohol :C=65*2; H=9*3.0'0947 gave 0.2362 CO, and 0.0831 H,O.C27H4302N,HN0, requires C = 68.0 ; H = 9.2 per cent.Solangustidine Sulphat e, (C,7H4,0,N)z,H,S04.-An alcoholicsolution of solangustidine was acidified with dilute sulphuric acid,when a white, amorphous powder separated.On boiling the mix-ture for some time the solid became crystalline, forming colourlessleaflets, which do not melt a t 330°, are very sparingly soluble inboiling alcohol, and insoluble in water :C=68*0; H=9*7.0.1938 * gave 0.0508 BaSO,. "SO, = 10.8.(C,7H4,02N)z,H,S0, requires "SO, = 10.4 per cent.* Dried at 130"568 TUTIN AND CLEWER:Solnngustidiiae p'crate was prepared by adding the requisiteamount of picric acid to a solution of the base in alcohol.Itformed yellow needles, which melted and decomposed a t 250O.Attempts were made to prepare the aurichloride and platini-chloride of solangustidine, but they resulted in failure, owing,apparently, to the fact that these derivatives are more readilysoluble than the corresponding hydrochloride.A cetylsolangustidiize, C,iH,202N*CO*CH,.-A quantity (2 grams)of solangustidine hydrochloride was boiled for one hour with aceticanhydride. When cool, the mixture was poured into ether, andthe ethereal solution shaken with aqueous sodium carbonate untilfree from acid. It was then dried and evaporated, and the residuecrystallised twice from ethyl acetate. Colourless, flattened needleswere then obtained, which melted a t 256O:0.1766,* on heating a t 130°, last 0.0032 H,O, which was quailti-0.0923 * gave 0-2530 CO, and 0-0820 HzO.0.3870 t in 21.38 of chloroform gave A t + 0'135O.C,,H,,O3N requires C = 76.5 ; H = 9.9 per cent.tatively reabsorbed on exposure to the air.H,O = 1-8.C=74.8; H=9*9.M.W. =491.0.0854 t ,, 0.2386 CO, ,, 0.0769 HZO. C=76.2; H=10.0.C,,H,,O,N,~H,O requires C = 75.0 ; H = 9.9 ; H20 = 1.9 per cent.M.W. = 455.The material conbained in the mother liquors from the acetyl-solangustidine was not homogeneous, and was found t o containunchanged solangustine. It yielded the pure acetyl derivative onmore prolonged acetylation.Attempts were made to prepare the hydrochloride of acetyl-solangustidine, and, although indications were obtained of theformation of such a compound, it was not found possible to isolateit. This was owing to the fact that the salts of the acetyl basedissociated with extreme readiness, and when crystallised fromalcohol, even in the presence of an excess of acid, they yield theoriginal acetyl compound.Acetylsolangustidine is remarkably stable towards alkalis, since,when boiled for several hours with alcoholic potassium hydroxide,i t is recovered unchanged.When, however, it is heated for a veryprolonged period with concentrated alcoholic potassium hydroxidei t slowly undergoes some change, but the amount of material avail-able was not sufficient to ascertain whether this resulted in theregeneration of solangustidine.When acetylsolangustidine is boiled for some hours with glacialacetic acid and concentrated hydrochloric acid, a yellow-coloured liquid is obtained, which exhibits a remarkably strong,* Air-dried substance..1- Dried at 130"THE CONSTITUENTS OF SOLANUhl ANGUSTIFOLIUM. 569fluorescence.of amorphous material.The reaction product, however, consisted onlyExamination of t h e Product (b).The product ( b ) (p. 563), which consisted of the material readilysoluble in cold, dry amyl alcohol, was of a dark brown, viscidnature, and nothing definite could be directly separated from it.It was dissolved in water, deprived of amyl akohol by means ofsteam, and then treated with sulphuric acid, when a quantity(2 grams) of rather impure solangustine sulphate was obtained,from which the pure salt could only with some difficulty beprepared.The acid filtrate from the crude sulphate, after being boiled,yielded to ether a small amount of quercetin. It was subsequentlymade alkaline with potassium hydroxide, and again boiled for sometime, when it yielded about 2 grams of 3 : 4-dihydroxycinnamicacid. The remainder of the product consisted only of resinousmaterial.The aqueous liquid (A), which had been extracted with amylalcohol, as above described, was kept for some time, when it gradu-ally deposited a further quantity (49 grams) of ruth.Afterfiltration, a portion of it (200 c.c.) was examined for the presenceof any further alkaloidal material. It was rendered alkaline withsodium carbonate, and extracted successively with ether, chloro-form, and amyl alcohol.The firstrmentioned solvent removed avery small amount of alkaloid, possessing a strong odour, Some-what resembling that of coniine, but the amount was insufficientf o r further examination. The chloroform and amyl alcoholremoved nothing.The remainder of the aqueous liquid (A), which amounted to4.5 litres, was concentrated somewhat under diminished pressure,and then treated with an excess of aqueous basic lead acetate.The resulting copious, yellow precipitate was collected and washed,after which it was suspended in water and decomposed by meansof hydrogen sulphide. The filtrate from the lead sulphide was thenevaporated under diminished pressure, when a dark brown, viscidproduct was obtained, from which nothing could be directlyseparated. This viscid, brown material was divided into two por-tions, one of which was heated for an hour with dilute sulphuricacid, but the resulting products were entirely amorphous.Theother portion was heated with aqueous potassium hydroxide in themanner previously described, when it yielded, in addition toamorphous products, a small amount of quercetin and a consider-able quantity of 3 : 4-dihydroxycinnamic acid570 TUTIN AND CLEWER:Isolation of 1-L4spuragir~e.A portion (about one-fifth) of the aqueous liquid (A), whichhad been treated with basic lead acetate, was slightly acidified withacetic acid, and then treated with a solution of mercuric nitratein dilute nitric acid until no further precipitate was produced.The precipitate was collected, washed, decomposed by means ofhydrogen sulphide, and the liquid filtered. The filtrate wasrendered slightly alkaline with ammonia, and then just acid withacetic acid, when it was concentrated under diminished pressureto a low bulk, The brown liquid so obtained was treated withanimal charcoal, after which, on keeping, it deposited a crystallinesubstance in the form of prisms. The latter was collected and re-crystallised from water, when it formed colourless prisms, meltingindefinitely a t 227-238O, and was identified as I-asparagine.Theamount obtained was about 0.25 gram, being equivalent to about0-02 per cent. of the air-dried plant'. (Found," H20=12*0.C,H,0,N2,H20 requires H20 = 12.0. Found,+ C = 36.5 ; H = 6.2.C,H,0,N2 requires C = 36.4 ; H = 6.1 per cent.)The remainder of the aqueous liquid (A) was deprived of leadby means of hydrogen sulphide, and concentrated under diminishedpressure t o a low bulk.The resulting syrup deposited no crystals,and no crystalline acetyl derivative could be prepared from it.It readily yielded d-phenylglucosazone (m. p. 212O), and thereforecontained a quantity of a sugar, probably consisting chiefly oflaevulose. It was carefully examined for the presence of rhamnose,but with a negative result.Examination of the Resilk ( B )The resin (B) was a soft, dark green mass, and amounted t o670 grams, being thus equivalent t o about 9.2 per cent. of theweight of the drug employed. It was dissolved in alcohol, mixedwith purified sawdust, and the thoroughly dried mixture extractedsuccessively in a large Soxhlet apparatus with light petroleum(b.p. 35-50°), ether, chloroform, ethyl acetate, and alcohol.Petroleum Extract of t?i,e Resilt.The petroleum extract of the resin was a dark green, fatty mass,and amounted to 540 grams. It was digested with about 2 litresof ether, and the mixture filtered, when a quantity (about 3 grams)* Air-dried substance. t Dried at 130"THE CONSTITUENTS OF SOLANUM ANGUSTIFOLIUM. 571of a solid was removed, which was examined in connexion with theurisaponifiable constituents.The ethereal filtrate was extracted with 10 per cent. hydrochloricacid, when a small quantity of a green, amorphous, alkaloidal pro-duct was removed. It possessed a strong, basic odour, but nothingcrystalline _could be obtained from it.The original ethereal solution was then extracted with aqueouspotassium hydroxide, when a quantity of a flocculent solidseparated. The latter was collected, when it proved to be thepotassium salt of a fatty acid, and was examined in connexion witha larger amount of a similar product obtained later.IsolcctioiL of n Pkytosteroliti, C33H5(i06.The alkaline, aqueous extract, which had been separated frointhe ether and solid material, as above described, was acidified andextracted with ether, when a quantity of a nearly black solidseparated, and was removed.Nothing crystalline could beobtained from it. The ethereal liquid was then shaken withaqueous potassium hydroxide, as before, when some neutralmaterial, which had been occluded during the first extraction withalkali, remained in the ether. The alkaline liquid was thenacidified and extracted with ether, during which operation aquantity (about 5 grams) of flocculent, green material separated,and was collected.This product was heated with acetic anhydride in the presenceof pyridine for half an hour, when, after concentration, colourlessleaflets separated.The latter substance, after recrystallisationfrom petroleum (b. p. 90-120O) and from alcohol, melted a t168-1 69O, and was identified as a tetra-acetylphytostrolin.(Found, C = 69.3 ; €I= 9.2.0.3885, made up to 20 C.C. with chloroform, gave a,-lOO’ in aA quantity of this acetyl derivative was hydrolysed by means ofalcoholic potassium hydroxide, when the resulting phytosterolin(phytosterol glucoside) was obtained.It separated from dilutepyridine in small, colourless crystals, melting a t 300O. (Found,C = 72.6 ; H = 10.5. Calc., C = 72.3 ; H = 10.2 per cent.)The benzoyl derivative was prepared by benzoylation in pyridinesolution. It crystallised from a mixture of chloroform and alcoholin colourless needles, melting a t 200O. (Found, C = 75-9 ; H = 7.8.Calc., C = 75.9 ; H = 7.5 per cent.)Calc., C = 68.7 ; H = 8.9 per cent.)2-dcm. tube, whence [a], - 25-7O.The ethereal solution containing the free, fatty acids, which hadbeen separated by filtration from the crude phytosterolin, as abov572 TUTIN AND CLEWER:described, was evaporated, and the dark green residue esterified bymeans of methyl alcohol and sulphuric acid.The resulting ester,dissolved in ether, was shaken with aqueous potassium hydroxide,when a considerable amount of dark green, phenolic resin was re-moved, from which nothing crystalline could be obtained. Theethereal solution was then dried and evaporated, when, after purify-ing the residual esters by distillation under diminished pressure,they were examined in connexion with a similar product obtainedfrom the combined acids, as described below.The ethereal solution of the neutral constibuents of the petroleumextract was evaporated, and the residue heated f o r two hours withan excess of alcoholic potassium hydroxide. Water was then added,and the mixture repeatedly extracted with ether.During thisoperation a quantity of a flocculent solid separated a t the junctureof the aqueous and ethereal layers. This was collected, when it wasfound t o consist of a mixture of the potassium salts of the higherfatty acids, and its examination will be described later.Zsola tion of Tria cont a n e, C,,H,, .The ethereal solution of the unsaponifiable material, which hadbeen separated from the alkaline, aqueous liquid and the potassiumsalt, as above described, was washed, dried, and evaporated. Theresidue, which amounted to 103 grams, was dissolved in alcohol,with the exception of a small amount of black, tarry material,which was discarded. On cooling the solution, a quantity (5 grams)of a solid separated, which was collected, and distilled underdiminished pressure.The distillate was crystallised twice f roniethyl acetate, when it formed lustrous, colourless leaflets, meltinga t 65'5O, and was identified as triacontane. (Found, C=85*1;H = 14.8. Calc., C =85.3 ; H = 14.7 per cent.)Isolation of a Phytosterol, C27H460.The alcoholic solution of the unsaponifiable material, from whichthe triacontane had been separated, was concentrated, and someethyl acetate and a little water added. On keeping this mixturefor some time a quantity (about 0.5 gram) of colourless leafletsseparated. This product had the properties of a phytosterol, andmelted a t 131°, but it did not appear to be homogeneous. It wasaccordingly converted into the acetyl derivative, which formedcolourless leaflets, melting a t 121°, and the latter crystallisedrepeatedly, both from alcohol and ethyl acetate.On regeneratingthe phytosterol and crystallising it, colourless plates were obtained,which were apparently homogeneous, and melted a t 134O THE CONSTITUENTS OF SOLANUM ANGIJSTIFOLIUM. 5730.2760," on heating a t 120°, lost 0.0138 H20.0,1029 I. gave 0.3160 C02 and 0.1120 H,O.H20=5*0.C=83.8; H=12'1.C,,H,,O,H,O requires H20 = 4.5 per cent.C2TH460 requires C = 83.9 ; H = 11.9 per cent.The liquid from which the crude phytosterol had been removedcontained a considerable quantity of a brown, sweebsmelling oil,from which no further crystalline material could be separated. Itwas examined for the presence of fatty alcohols by the phtlialicanhydride treatment, but with a negative result.Emtiiiticrtioit of t h e Fatty Acids.The nlkaliiie liquid from which the unsaponifiable materialhad been removed by means of ether was acidified and distilIedwith steam.This removed a small amount of a volatile acid havingan odour of valeric acid. The mixture was then extracted withether, when a small amount of phytosterolin separated, and wasremoved. The ethereal liquid was then dried and evaporated, andthe dark green residue esterified by means of methyl alcohol andsulphuric acid. The resulting methyl esters were then freed froma little unchanged acid and much chlorophyll by shaking theirethereal solution with aqueous potassium hydroxide, and then wash-ing i t with water, after which the liquid was dried and evaporated.The residue was then purified by distillation under diminishedpressure, when a quantity (94 grams) of methyl ester was obtained.This product was added to the esters of the free acids previouslymentioned, which amounted t o 118 grams, and the whole hydrolyseclby means of alcoholic potassium hydroxide.The resulting acidswere isolated, and separated into their saturated and unsaturatedcomponents by means of the lead salt, in the usual manner.The Unsaturated Acids.-These acids were converted into themethyl ester, which amounted to 163 grams, and the latter distilledseveral times under diminished pressure, when the following frac-tions were collected: i, Below 215O (16-5 grams); ii, 215-218O(33.9 grams); iii, 218-222O (73-0 grams); iv, 222-225O(16-2 grams)/20 mm.The iodine values of fractions i, ii, iii, and iv were respectively182.5, 200.7, 210.9, and 207.3.Fraction iii, on analysis, gaveC = 77.5 ; H = 11.2 per cent. (Methyl linolate requires C = 77.5 ;H = 11.5 per cent. ; I.V. = 172.7 ; and methyl linolenate requiresC = 78.1 ; H = 10.8 per cent. ; I.V. = 261.)It would therefore appear that the unsaturated acids consistedessentially of linolic and linoleiiic acids.* Air-dried substance. t Dried at 110"574 TUTIN AKD CLEWER:?‘he Satirrccted Futty Acids.-These acids amounted to 40 grams.They were mixed with the fatty acid (6 grams) obtained from thepreviously mentioned sparingly soluble potassium salts, and con-verted into the methyl ester.The latter was fractionally distilledseveral times under diminished pressure, when the following frac-tions were collected : i, Below 200O; ii, 200-205°; iii, 205-210O;iv, 210-215°; v, 215-225O; vi, 225-235O; vii, 235O+ /a0 mm.Fractions i and ii solidified on cooling, and were found to consistof methyl palmitate. They yielded palmitic acid, melting a t 63O.(Found, C=74.9; H=12*5.Fractions iii, iv, and v all yielded impure products, whichappeared to consist of mixtures of palmitic and stearic acids.(Found, C=75.6; H=12.6. Calc. for C,,H320,, C=75*0; H=12.5;for C,,H,,O,, C=76*1; H=12-7 per cent.)Calc., C=75*0; H=12.5 per cent.)Isolation of Cluytziiuic Acid, C,,H4,0,.The above-mentioned fraction vi of the methyl esters partlysolidified.On hydrolysis it yielded an acid, which was crystallisedmany times from alcohol, from acetic acid, and from ethyl acetate,when it melted constantly a t 69O. On being compared directlywith cluytinic acid obtained from C’luyt~ia sirnilis (T,, 1912, 101,2226) and from hops (T., 1913, 103, 1283), it was found t o beidentical with both these preparations. I n order further to estab-lish its identity, however, it was converted into its methyl ester.The latter formed colourless needles, melting a t 47O, and wasidentical with the corresponding derivative prepared from thecluytinic acid of hops (loc. cit.). (Found, C=77.5; H=13.1.C22H4402 requires C = 77.6 ; H = 12.9 per cent.)Fraction vii of the esters, together with the residues of highboiling point which remained in the flasks after the fractionaldistillations, was redistilled, when it solidified on cooling. Oncrystallisation from ethyl acetate it yielded methyl cerotate, meltring a t 60°.On hydrolysis, the latter gave cerotic acid, melting a t78’5O. Calc., C = 79.0 ; H = 1 3 2 percent.)(Found, C= 78.9 ; H = 13.2.Ethereal Extract of the Resitt.This extract of the resin was dark green, and amounted t o46 grams. A portion of it (3 grams) was very sparingly soluble inether, and formed a nearly black powder. The latter was extractedin a Soxhlet apparatus for a short time with ethyl acetate, whenabout 1 gram of crude phytosterolin remained undissolved. Theethyl acetate extract was evaporated, and the residue distilled underdiminished pressure, when, on crystallisation from ethyl acetate, iTHE CONSTITUENTS OF SOLAKUM ANGUSTIFOLIUM.575formed colourless leaflets, meltiiig a t 85O. This product, togetherwith a further amount of similar material, obtained as describedbelow, was found t o consist of a higher fatty acid. It was con-verted into the methyl ester, which formed colourless leaflets, melt-ing constan6ly a t 71O. On hydrolysing the latter, the original acidwas regenerated, and, on crystallisation from acetic acid, its melt-ing point remained unchanged, a t 8 5 O :0.1032 gave 0.3001 CO, and 0.1242 H20. C=79*3; H=13-4.0.0602 ,, 0.1748 CO, ,, 0'0715 H20. C=79*2; H=13.3.C2,H,02 requires C = 79.2 ; H = 13.2 per cent.CNH6002 ,, C=79.6; H=13*2 per cent.It would thus appear that this higher fatty acid was eithermelissic acid or a lower homologue of the latter.The low meltingpoints of the acid and its ester would point to the latter conclusionbeing the more correct.Ths ethereal solution of the more readily soluble portion of theethereal extract of the resin yielded a small amount of green,amorphous, alkaloidal material when extracted with hydrochloricacid. On subsequently shaking it with aqueous sodium carbonatenothing was removed. When, however, the ethereal solution wasshaken with aqueous potassium hydroxide, the greater portion ofthe dissolved material was removed from the ether, a portion of itforming an insoluble potassium salt, which was collected. Thelatter proved to be the salt of the above-described higher fatty acid.The material dissolved by the alkali, which represented thegreater part of the ether extract of the resin, consisted of a darkgreen, phenolic resin, from which nothing definite could beobtained.Chloroform, Ethyl Acetate, and Ak0h01 Extracts of the Resin.The chloroform, ethyl acetate, and alcohol extracts of the resinamounted to 36, 7, and 33 grams respectively.They all consistedof dark green, amorphous resins, from which nothing definite couldbe obtained, with the exception of a small amount of solangustine.The latter was isolated in the form of its sulphate from the alcoholextract of the resin.Summary and Physiological Tests.The material employed in this investigation consisted of theleaves, twigs, and flowers of Solanum angustifolium, Ruiz et Pavon,which had been obtained from Peru,For the purpose of a complete examination, 30.92 kilograms o576 MORRELL AN b BURGEN :the dried inaterial were employed. This material was ground,completely extracted with hot alcoliol, and the resulting extractdistilled in a current of steam.From the portion of the extract which was soluble in water therewere isolated the following substances : (i) Quercetin ; (ii) rutin,C;7H300,6,3H20 ; (iii) I-asparagine ; (iv) a new gluco-alkaloid,solungus tine, C,,H,O,N,H,O. On hydrolysis, solangustine yields~solangustidine, C27H430$T, together with 1 molecule of dextrose.The aqueous liquid also contained small amounts of amorphous,alkaloidal material, and a considerable quantity of a sugar, whichapparently was lzvulose, together with viscid, amorphous products.Some of the latter yielded quercetin and 3 : 4-dihydroxycinnamicacid on treatment with alkalis.The portion of the original extract which was insoluble in wateryielded, in addition to much chlorophyll and resinous material, thefollowing compounds : (i) Triacontane, c30HG2 ; (ii) a phytosterol,c&4@ ; (iii) a phytosterolin (phytosterol glucoside), C33H560ri ;(iv) palmitic, stearic, cluytinic, and cerotic acids, together with amixture of linolic and linoleiiic acids. It furthermore gave a smallamount of the above-mentioned new gluco-alkaloid, solangustine,and a higher fatty acid, which was either melissic acid, C",OHGOO,,or a lower homologue, C,,H,,O,.The following physiological tests were conducted a t the Well-come Physiological Research Laboratories by Dr. H. H. Dale, towhom our thanks are due.An amount of the total alcoholic extract, equivalent to 3.5 gramsof the drug, and 0.48 gram of solangustine, were separately adminis-tered to a dog, but no perceptible effect of any kind resulted. Theamorphous alkaloidal material, which occurred to a small extent i nthe plant, yielded a similarly negative result.THE WELLCOME CHEMICAL RES~M:CH LABORATORIES,LONDON, E.C