1518 JOKETT ON ATISINE, XCIX.-Cont~.i~utio.lis to ouq' Knowledge of the Aconite Alkaloids. Part XIII. On Atisine, the Alkaloid o f Aconitum heterophyllum. By HOOPER ALBERT DICKIMON JOWETT, L).Sc. Lond., Research Fellow of the Pharmaceutical Society. AMONGST the many species of Aconitum known to botanists, t<here is one-Aconitzcm hetem@ylZum-which especially attracts attention, because it is not poisonous. I n this respect it differs from the other species of Aconitum which have been chemically examined. The roots of this plant are sold in the Indian bazaars as utis, or atees, and have long been used in India on account of their medicinal value, b u t very little is known concerning the chemical constituents of the root. I n 1873, Broughton (Blue Book, h'ast India Cinchona Cultivation, 1877, 133) isolated an nlkalojid from the roots of t,he plant, to which 'he gave the name atisine, aid proposed for it the formula C46H,,N,0,. The nlkaloi'd was only obtained as a n amorphous varnish, yielding, however, a crystalline sulphate and hydrochloride.The platini- chloride was also prepared, and the formula above given was deduced from the analyses of this salt, which, however, have not been published. Tn 1679, Wasowicz (Arch. Pliarm., 214, 193) isolated the same alkaloi'd from the roots of the plant, and also aconitic acid. Hc adopted Broughton's formula for atisine with very slight! modifica- tion, and prepared the halojid salts in the crystalline condition. The hydriodide was found to be only sparingly soluble in cold water, arid could thus be easily purified by recrystallisation ; on analysis, results were obtained corresponding with the formula, CP~H,,N~O~,HI, H20.In 1879, Dr. Alder Wright (Year Book of Phamaacy, 1879, 422) proposed for atisine the formula, C22H,IN02, based on a single ana- lysis of the amorphous aurichloride, the quantity at his disposal being too small for further examination.THE ALKALOID OF ACONLTUM HETEROPHYLLUM. 1519 In the course of the investigation described i n this paper, the properties of the base and its salts have been carefully determined. A satisfactory method has been devised for the extraction of the roots and the preparation of pure material, and the presence of aconitic acid, mentioned by Wasowicz, has been confirmed. Extr*actioiL of the Alkaloid fyom the Root, and its PzcriJicatioia.The solvents, previously used to extract the alkaloXdal salts from the aconite roots, hare been chiefly alcohol acidified with 1 per cent. tartaric acid, or more recently amylic alcohol (steam-dist,illed fusel oil). After a number of experiments with different solvents such as light petroleum, ether, alcohol, &c., had been made, it was found that the most satisfactory results were obtained by using a mixture of methylic arid amylic alcohols, in the proportion of three volumes of methylic alcohol to one volume of arnylic alcohol, the commercial wood spirit and fusel oil being employed for this purpose. The advantages of this solvent for the extraction of the roots are (i) the roots are completely extracted in a very short time, and the percolate does not contain very much colonring matter, resin, $c., although more than if amylic alcohol alone were used ; (ii) the percolate can be reduced to a small bulk without very much heating, as the methylic alcohol can be distilled off at a low temperature under diminished pressure, thus prevenhg decomposition of the alkaloid by heat ; (iii) the alcoholic solution thus obtained deposits a large quantity of fat and colouring matter, from which the alcohol containing the alkala'id in solution can be separated by decantation.The alkalo'id can then be extracted by shaking the alcoholic solution with 1 per cent. aqueous sulphuric acid, and following the methods described when amylic alcohol alone is used. This solvent, which has since been successfully employed for the extraction of other species of aconite roots, possesses the advantage of both methylic and amylic alcohols with none of their attendant disadvantages.About 60 lbs. of the root in fine powder were extracted by per- colation with t'lie above medium, and the percolate obtained, which was of a daiak brown colour, was distilled over a water bath, when most of the methylic alcohol, together with a little amylic alcohol, was removed. On standing, the alcoholic solution deposited a large quantity of very dark coloured fat, which was separated by decanta- tion; the solution in amylic alcohol mas then repeatedly shaken with 1 per cent. aqueous sulphuric acid until no more alkaloid was extracted, the acid liquid shaken twice with chloroform to remove aniylic alcohol, and then, after neutralisation, concentrated by evaporation on a water bath.The concentrated solution, which was7 520 JOWETT ON ATISINE, dark coloured, was rendered alkaline by aqueous sodium hydroxide, and then repeatedly shaken with ether or chloroform until no ap- preciable amount of alkalo'id mas any longer extracted. It is not possible, however, by this method to extract the whole of the alkaloid from the aqueous liquid. The ethereal (or chloroform) solution, after being washed once or twice with water, is distilled, and the residue dissolved in dilnte sulphuric acid, using as small an excess as possible. The solution thus obtained is dark brown, and still contains in addition t o the alkaloidal salt, a large quantity of resinous aiid colouriDg matter, which prevents crystallisation when the solution is evaporated, These attendant impurities are best removed by fractionally pre- cipitating the alkaloid with aqueous sodium hydroxide, the first fractions containing nearly the whole of the impurities, whilst the later fractions consist of almost pure alkalo'id ; this fractionation is repeated several times, until the acid solution of the alkaloid is almost colourless.The freshly precipitated alkalo'id, after being col- lected and washed, is suspended in water and dissolved in exactly the right quantity of dilute hydrochloric acid to form a neutral solution of atisine hydrochloride. This, when slowly evaporated on a water bath, yields crystals, which are collected and re-crystallised once 01' twice from water, and then from alcohol ; on adding ether to tho alcoholic solution until there is a faint permanent turbidity, and then setting it aside in a stoppered bottle, fine needle-shaped crystals of the pure hydrochloride are obtained.These are collected and dried in the usual manner. For the extraction of the remaining traces of alkaloid in the different filtrates obtained during fractionation, and in the mother liquor after extraction with ether, a different method is employed. The solution is rendered faintly acid and excess of po tassio-mercuric iodide added, the precipitate thus produced is first thoroughly washed several times by decantation with water, and then suspended in water, and decomposed by passing hydrogen sulphide through the mixture ; the precipitate is then collected, extracted by boiling water until no more is dissolved, and the aqueous extracts mixed and allowed to cool ; on standing, very faintly coloured crystals of atisine hydriodide separate, which may be purified by recry stallisation from hot water.A few crystals can be obtained by concentrating the original filtrate, but owing to the sparing solubility of the salt in cold water, scarcely anything is left in the mother liquors. In this way the whole of the alkaloid is obtained as the sparingly soluble crystalline hydriodide, and the easily soluble hydrochloride, the amount of hydriodide contained in the mother liquor being so small that it may be neglected.THE ALKALOIII OF ACONITUM HETEROPHYLLUM. 1521 The total amount of alkaloyd contained in the root is small, from 0.2 to 0.3 per cent., and no evidence was obtained of the existence of any alkalo'id other than atisine.IdentiJication of Aconitic acid from the Root. As the melting point of the acid obtained by Wasowice did not agree with that recorded for aconitic acid by other observers, I pre- pared the acid from the root, and examined its properties. It was isolated by means of its lead salt, which was then decomposed by hydrogen sulphide, and purified by crystallisation from water until its melting point was constant. It is a white, ci*ystalline substance, very soluble in ether or water, and, when dried at loo', melted and decomposed with effervescence at 191.5' (corr.). Aconitic acid melts, and decomposes with effervescence, at 191'.On analysis, the acid and its silver salt yielded the following data. 0.0382 acid required 12.5 C.C. N/20 soda, 0.0758 silver salt gave 0.0494 Ag. Ag = 65.17. Calculated for CSH606, 13.1 C.C. and 63.45 per cent. silver. The acid is thus proved to be aconitic acid. Composition and Properties of Atisine and its Salts. The base, which has ouly been obtained as a colourless varnish, is prepared as follows. The aqueous solution of a salt (preferably the hydrochloride) is rendered alkaline by sodium hydroxide, and the alkaloid, which is precipitated in white flocks, is extracted by shaking with ether ; the ethereal solution, after being shaken with a little water until free from alkali, is dried with calcium chloride, and the ether removed by distillation. The base is thus left as an amorphous, colourless varnish, which is slightly soluble in water, freely soluble in alcohol, ether, or chloroform, and insoluble in light petroleum.It readily undergoes decomposition when heated, becoming brown, and forming a resinous substance. All attempts to crystallise i t from various solvents failed. Its alcoholic solution is laevorotatory, although the salts are dextrorotatory, a determina- tion yielding the following results. a[19'] = - 2.64"; 1 = 2 dm; c = 6.728. Whence Atisine hydrochloride, C22HslN02,H C1, is prepared by dissolving the base in exactly the requisite quantity of aqueous hydrochloric acid, and crystallising the salt from the aqueous solution thus obtained. It may also be crystallised from alcohol by the addition of ether. As VOL.LXIX. 5 K1522 JOWETT ON ATISINE, thus obtained, the pure hydrochloride forms long, transparent prisms, occurring either singly or in rosettes. The crystals are anhydrous, as there is no loss in weight when they are heated at 100-130" for five hours. The salt is very soluble in water and alcohol, less so in benzene and acetone, and sparingly in dry chloro- form ; it is insoluble in ether and in light petroleum. When dried at loo", it melts and decomposes with effervescence at about 296O (corr.). The aqueous solution is dextroro tatory, two determinations yielding the following results : (i) a[19"] = +0*61"; I = 2dm; c = 1.691. Whence Whence (ii) a[24'] = +0*38"; I = 2dm; c = 1.002. loo 0*38 = [a],,.= +18.90. 2 x 1.002 Mean of two determinations.On analysis- 0.1776 of salt gave 0.0684 AgC1. The molecular weight of the salt, determined by the depression of [a]= = +18.46'. C1 = 9.53. C22H3J",,HC1 requires C1 = 9.38 per cent. the freezing point of glacial acetic acid, gave the following result: Weight of salt taken.. . . .. . . Weight of solvent.. . . . . . . . . Depression of freezing point. . = 0.16'76 = 5.327 = 0.34" 3.146 0.34 39 x - = M = 360.8. C,H3,NO2,HC1 requires 377.4. Atisine hydrobromide, C22H,,N0,,HBr, prepared in a similar manner to the hydrochloride, crystallises from a mixture of alcohol and ether in long needles, which are anhydrous. It is freely soluble in water or alcohol, less so in chloroform, insoluble in ether, and melts and decomposes with effervescence at 273" (corr.). The aqueous solution of the salt is dextrorotatory, a determination giving the following results : a[15"] = +0*386"; 1 = 2dm; c = 0.7915.Whence Whence .A tisine hydkodide, C22H,,N O,,HI, cannot be obtained by the action of aqueous hydriodic acid on the base, since iodine is liberated, and a coloured substance--probably an iodo-derivative-formed. It is,THE ALKALOID OF ACONITUM HETEROPHYLLUM. 1523 however, very easily prepared by adding Mayer's reagent to a solu- tion of a salt of atisine, and treating the precipitate as before described. Atisine hydriodide is also precipitated on adding aqueous potassium iodide to a strong aqueous solution of an atisine salt. The hydriodide is best purified by dissolving it in boiling water, when, on cooling, an abundant deposit of crystals is obtained.It may also be crystallised by the spontaneous evaporation of an alcoholic solution, or by the addition of ether to the solution. The crystals are anhydrous, transparent plates or tables, occurring singly or in rosettes ; they are very sparingly soluble i n cold, more so in hot water, sparingly soluble in alcohol or acetone, and almost in- soluble in chloroform, ether, light petroleum, or benzene. A determination of the solubility of this salt in water gave the following results : I. 6.9924 solution at 90" gave 0.0926 salt. 11. 8.7992 solution a t 15O gave 0.0182 salt. Hence 1.3 grams of salt dissolves in 100 grams of water at 90". Hence 0.2 -"ram of salt dissolves in 100 grams of water at 15". The pure salt, when dried at loo", melts and decomposes with effervescence at 279-280" (corn.).The aqueous solution of the salt is dextrorotatory, two determinations yielding the following re- sults : (i) a[19'] = + 0.19"; I = 4dm ; c: = 0.1808. Whence (ii) a[19'] = +0.lo; I = 2dm; c = 0.1'74. Whence loo *.' = [a]= = +28*7'. 2 x 0.174 Mean of two determinations [a]= = +2'7*4'. On analysis, the salt furnished the following results : 0.1652 gave 0.3468 GO2 and 0.1086 H,O. C = 57-02 ; H = 7.30. 0.0914 ,, 0.1904 GO, ,, 0.0598 H20. C = 56.77; H = 7-28, 0.1926 ,, 0.097 AgI. I = 27.21. 0.0926 ,, 0.0462 AgI. I = 27.02. 0.2636 ,, 0,1312 AgI. I = 26.89. 0.3753 ,, 11 C.C. moist nitrogen at 13" and 755.4mm. N = 3.44 C22H31N02,HI requires C = 56-29 ; H = 6-82 ; N = 2.99; I = 27.07 per cent. Atisine nitrate, CZ2H3,NO2,HNO3, is best obtained by adding a slight excess of aqueous silver nitrate to a hot aqueous solution of the pure hydriodide, filtering off the silver iodide, carefully removing the excess of silver nit-rate by dilute aqueous hydrochloric acid, filtering1524 JOWETT ON ATISINE, off the precipitated silver c h i d e , neutralisiog the filtrate with dilute ammonia, and concentrating the aqueous solution on the water bath.On cooling, the salt crystallises out in very well-defined hexagonal plates, which are best purified by recrystallisation from alcohol and ether. The cryRtals, which may occur singly or in rosettes, are, like the other salts of atisine, anhydrous ; they are soluble in water, particularly in hot water, fairly so in benzene or acetone, more so in alcohol, sparingly in chloroform, and insoluble in ether or light petroleum.The dried salt melts sharply at 252' (corr.). The aqueous solution is dextrorotatory, a determination giving the following result : a[i80~ = + 0.580; I = 2am; = 1.0255. Whence loo 2 x 1.0255 o'58 = [a]" = +28*3O. On aualysis, the following data, were obtained. 0.144 gave 03458 GO2 and 0.1032 H20. C = 65.48 ; H = 7.96. 0.182 ,, 0.4368 ,, ,, 0'1302 ,, C = 65-43; H = 7-94. 0,1588 ,, 9.5 C.C. moist nitrogen at 13Oand 761.5 mm. N = 7.08, CzzH31N02,HN03 requires C = 65.34 ; H = 7.92 ; N = 6-93 p. c. In the determination of the nitrogen contained in the hydriodide and nitrate, a difficulty was encountered similar to that described by Dnnstan and Carr in the case of aconitine (Proc., 1896, 48), and concordant results were only obtained by mixing the salt with pow- dered copper oxide and lead chromate, and conducting the combustion in a vacuum.Atisine aurichloride, Cz2R3,NO2,HAuCl4, prepared by the usual methods, could only be obtained as an amorphous powder, soluble in alcohol, but insoluble in ether or light petroleum. 0.0682 gold salt, on ignition, gapve 0*0202 Au. C22H31N02,HAnClo requires Au = 28.82 per cent. Atisine platinichloride, ( C22H81N02),,HzPtC16, is a yellow crystalline powder, sparingly soluble in cold, more so in hot, water, and melts sharply at 229O (corr.) with effervescence. Au = 29.61. 0.0609 gave 0.011 Pt. The result of these analyses of pure material confirm conclusively the formula for atisine (CZzH3,NO2) proposed by Wright. When atisine is treated with alkalis or acids under varying con- ditions, no acid is eliminated, but a new base is formed.To prepare the latter, atisine i n alcoholic solution is heated with alkali for some houra, either in a sealed tube or in a reflux apparatus ; the liquid is then neutralised with aqueous sulphuric acid, the alcohol removed P t = 18.06. (Cz2H~1NO2),,H2PtCI, requires Pt = 17.78 per cent.THE ALEALOID OF ACONITUM HETEROPHYLLUM. 1525 by distillation, the solution rendered alkaline with aqueous sodium hydroxide, and extracted with ether. On distilling off the ether, a non-crystalline, colourless varnish is left ; none of the salts could be obtained crystalline, but the anrichloride and platinichloride were prepared and analysed, and the figures obtained on analysis agree best for a base represented by the formula C22H3,N02,H,0.0.171 gold salt gave 0.239 CO, and 0.0784 HzO. C = 38-01 ; H = 5.08. 0.1026 gave on ignition 0.0292 An. An = 28.4. Two separate fractions (I and 11) gave on analysis : Au = 28-06. 11. 0.0370 ,, 0.0104 ,, Au = 28.10. Au = 28.08 per cent. The platinichloride, which melts at 236' (con-.), was analysed : 0.0536 gave 0.0094 Pt. Pt = 17.53. (C,,H,,NO,,H,O),,H,PtCJ, requires Pt = 17-21 per cent. The same bnse was obtained when atisine was heated in aqueous solution with dilute sulphnric acid, the analysis of the platinichloride thus formed giving the following result : I. 0.0588 gave 0.0165 Au. C~H,N02,H,0,HAuC14 requires C = 37-82 ; EI = 4-87 ; 0.0298 salt gave 0.0051 Pt. Calculated, Pt = 17.21 per cent.When the hydriodide is heated with fuming hydriodic acid, no methylic iodide is formed, thus proving the absence of methoxyl groups in the molecule. When atisine hydriodide is heated with soda lime, a white sub- stance sublimes which has the odour of a cresol ; this is insoluble in cold water, slightly soluble in hot water, but easily in alcohol 01' ether. Lack of material prevented the further examination of this decomposit,ion product. Professor Cash, of Aberdeen, who has examined the physiological action of atisine nitrate, reports that in small doses it is not toxic, and that its action in some respects resembles that of aconine. Pt = 17.11. The results of this investigation show that atisine does not present any close analogy to the alkalo'ids of the other well known species of aconite ( A . napellus, ferox, japonicum). The molecule is apparently less complex and much more stable, and since it yields a hydrate when heated with alkalis or acids, i t might be inferred that i t is the anhydride of atisine monhydrate ( C22H,3N03). There would thus be a relationship between these bases similar to that existing between aconine and pyraconine. The material used in ihis investigation was collected in India, at FOL. LXIX. 5 L1526 DOUCJAL: EFFECT O F HEAT the instance of the Scientific Department of the Imperial Institute, a,nd the inquiry has been conducted in the Research Laboratory of the Pharmaceutical Society. In conclusion, I desire t o express my warmest thanks to Professor Dunstan, both for having suggested the research, and for the very kind encouragement given to me during the progress of the work.