82 MCPENZIE AND STRATHERN REACTIONS OF XIV.-Reactions of Displacement in the Tropic Acid Group. Part I. By ALEX. MCKENZIE and ROBERT CAMPBELL STRATHERN. THE proof of the constitution of a-aminohydratropic acid, NH,-CPhMe*CO,H was afforded by Tiemann and Kohler (Ber., 1881 14 1980) by the synthesis from acetophenonecyanohydrin and alcoholic ammonia. McKenzie and Clough (J. 1912 162, 390) in dealing with the Walden inversion described the resolution of the r-acid into its optically active components. The study of the isomeric p-acid NH,*CH,*CHPh*CO,H has not so far been advanced to the corresponding stage. For one reason this acid has no particular interest with relationship to the Walden inversion the amino-group not being directly attached to the asymmetric carbon atom.The identification of the acid has also been attended with some confusion. By the action of aqueous ammonia on p-bromohydratropic acid Merling (AnnaZen, 1881 209 11) obtained crystals m. p. 169-170" (compare also Fittig and Wurster ibid. 1879 195 158). Assuming that the reaction proceeded on the lines CH,Br*CHPh*CO,H + NIP,*CH,*CHPh*CO,H, Merling concluded that he had to deal with p-aminohydratropic acid and the analytical evidence supported this assumption DISPLACEMENT IN THE TROPIC ACID GROUP. PaRT I. 83 Posner however in continuation of his work on the action of hydroxylamine on unsaturated acids obtained from atropic acid an amino-acid which melted and decomposed a t 234" (Ber. 1903, 36 4315; Annulen 1912 389 109). All the evidence pointed to this substance being the true p-aminohydratropic acid.Since Posner (Ber. 1905 38 2316) had shown that the action of ammonia on p-bromo-p-phcnylpropionic acid leads to the formation not of p-amino- p-phenylpropionic acid but of p-hydroxy-p-phenylpro-pionamide he suggested that P-bromohydratropic acid might perhaps exhibit a similar behaviour and that Merling's product was not the amino-acid a t all but was the isomeric tropamide. The latter suggestion has been borne out. In the present paper it is show1 that r-p-chlorohydratropic acid is converted into r-tropamide (m. p. 170-171") by the action of aqueous ammonia. The amide was also prepared directly from ammonia and methyl dl-tropate; it has the same melting point and it possesses the same properties as the compound described by Merling.We have also examined the action of hydroxylamine on atropic acid and find that Posner's p-amino-acid is not the sole product, and that the oxime of phenylacetaldehyde is also formed. It is, however rather curious that the formation of tropic acid by the action of nitrous acid on the amino-acid proceeds in a somewhat irregular manner. The preparation of Z- p-chlorohydratropic acid is described the resolution of the r-acid being effected in methyl-alcoholic solution with morphine as the resolving alkaloid. When hydrolysed the Z-acid which has [.ID - 122.6" in benzene gave a lzvorotatory tropic acid with [ R I D - 58.1" (c = 2.4348) in ethyl-alcoholic solution whereas the value for the optically pure acid is [ R ] ~ - 72.5" (c = 2.578) in the same solvent (NcKenzie and Wood, J.1919 115 828). Although the chlorine is not directly attached to the asymmetric carbon atom some racemisation had apparently taken place during the substitution of the hydroxy-group for the chlorine. In the light of recent work in a different field (McKenzie and Roger J. 1924 125 214S) the action of ammonia on the E-chloro-acid was of some interest and we have therefore examined it for the purpose of finding if optical activity would persist in the trans-formation into the amide. Experiments showed that a levorotatory tropamide did actually result. Under the conditions employed, the rotation of the product was less than that of the amide obtained by the direct action of ammonia on ethyl I-tropate. This formation of an optically active tropamide has led us t 84 MCKENZIE AND STRATHERN REACTIONS OF conclude that a p-lactone is produced as an intermediate phase, thus : Th Ph Z-H-$-CH,Cl -+ Z-H-$+-CH,Cl + C0,H CO,*NH, Ph Fh Z-H-Y-CH2 -+ Z-H-Y-CH,*OH.cod CO*NH,. Although on those lines the actions do not occur by substitution of groups directly attached to the asymmetric carbon atom some racemisation accompanied the change. This was not altogether unexpected since in tropamide we have the phenyl hydrogen, and carbonyl groups all directly attached to the asymmetric carbon atom a system which a t any rate in the presence of alcoholic alkali is particularly prone to racemisation under certain conditions (McKenzie and Smith J. 1922 121 1348). In the field of optical activity it has been shown by Holmberg (ArEiv Kern.Min. CeoZ. 1917,6 No. 23 pp. 1-33 and earlier papers) that the formation of a p-lactone is a stage in the interconversion of d- and I-malic acids by the Walden inversion. Moreover Z-iodo-succinic acid when treated with ammoniacal silver nitrate gives a mixture of fumaric acid and d-p-malamic acid the latter being formed by the action of ammonia on the d-malo-lactone (compare Walden and Lutz Ber. 1897 30 2796; Lutz Ber. 1902 35, 2460 4369; 1908 41 841). Reference has already been made to the fact that Posner obtained r- p-hydroxy- p-phenylpropionamide by the action of ammonia on r- p-bromo- p-phenylpropionic acid. It seems reasonable to suppose that this action proceeds on similar lines to that of ammonia on P-chlorohydratropic acid : ph p Ph Vh H-TBr -+ H T - B r CH,*CO,H CH2*C0,*NH4 CH,*CO CH,*CO*NH,.It will be observed that the p-lactone which is suggested as an intermediate phase is isomeric with the one postulated in the scheme with p-chlorohydratropic acid. Arising from the results recorded in the present paper and more especially from former work by one of us it is desirable that Posner’s observation should be extended to the optically active p-bromo-p-phenylpropionic acids so as to enable the following scheme to be completed: -+ HT-? -+H-$+-O DISPLACEMENT m THE TROPIC ACID GROUP. PART I. 85 dext,ro- dextro-1 zvo - C,H5*CHBr*CH,*C0,H dextro - C GH ,*CH( OH) *CH,*CO*NH, (McKenzie and Humphries J. 1910 97 121; McKenzie and Martin J. 1913 103 112; McKenzie Martin and Rule J., 1014 105 1583; McKenzie and Smith J.1922 121 1348). Owing to the announcement in the current number of the Pro-ceedings (1924 p. 77) of a paper by Seiiter and Ward entitled " Studies on the Walden iiiversion. Part IX. The influence of the solvent on the sign of the product in the conversion of 3-bromo-p-phenylpropioiiic acids to 3- hydroxy- 9 -phenylpropionamides " TF'C are not meanwhile invcstigatiiig this matter further so far as the action of ammonia on the optically active $-bromo-F-phenyl-propionic acids is concerned. At present we are engaged on a further study of the action of nitrous acid on r-p-aminohydratropic acid and also propose to undertake the resolution of this acid with the object of examining the resulting optically active acids.E X P E R I Jr E 3 T A L. Action of Hydroxyhnine o n Atropic Acid.-Posner (Anncrieih, 1912 389 112) prepared r-$aminohydratropic acid by heating atropic acid with an alcoholic solution of hydroxylsmine for 1 hour. By analogy with his previous work (Ber. 190.5 38 2316; compare also McKenzie and Tuclhope J. 1924,125,928) i t might be expected that the formation of the amino-acid would be accompanied by that of phenylncetaldoxiin~.. Kydroxylamine hydrochloride (25 g.) in water (25 c.c.) was adclcd to a solution of sodium (8.5 g.) in ethyl alcohol (280 c.c.), and t hc precipitated sodium chloride removed. Atropic acid (2.3 g.) prepared by the dehydration of atrolactinic acid (McKenzie and Wood Zoc. cit.) was added and the solution boiled for 13 hours.Ammonium carbonate was deposited in the condenser tube. -4fter 3 clays in the ice-chest the amino-acid (8 g.) was collected; it was free from the hydroxylamiiio-acid sirice it did not reduce either Fehling's solution or ainrnoniacal silver nitrate ; i t melted and decomposed a t 222-224" (Posner gives 233"). An additional 0.5 g. was obtained by boiling the filtrate for 2& hours longer. After distillation of the bulk of the alcohol an exccss of water precipitated 8 g. of crude phenylaceta!dosime which crystallised from water several times gave the pure substance in lwtrous needles m. p. 103-104" (Found C = 70.9; H = 6.8. Calc., C = 71.1 ; H = 6.7%). Bouvcault and M7ahl (C'ompt. rend. 1002, 134 1147) gave m. p. 103". This osime was isolat cd as follows 86 MCKENZIE AND STRATHERN REACTIONS OF r- P-Aminohydratropic acid is sparingly soluble in water benzene, light petroleum acetone chloroform carbon disulphide and ether.Posner (Ber. 1903 36 4315) described this acid as a-amino-a-phenylpropionic acid (a-aminohydratropic acid) ; its solution in hydrochloric acid was acted on by nitrous acid to give an acid, m. p. 89-90" which was supposed to be atrolactinic acid. Posner was however working with small quantities of material and was misled by the erroneous data of other authors. In a subsequent paper (Annulen Zoc. cit.) the correct constitution of the amino-acid is given. The product which Posner obtained by the action of nitrous acid was apparently impure tropic acid. We have repeated this action several times and find that it proceeds by no means smoothly.The gradual addition of potassium nitrite to the solution of the amino-acid in dilute hydrochloric acid caused the deposition of an oil which was certainly not tropic acid nor did it appear to be a-isotropic acid since it showed no signs of solidifying when nucleated with the latter acid. The aqueous solution was separated from the oil and extracted with ether. By crystallising the resulting acid several times from benzene r-tropic acid was isolated in very small yield (0.1 g. of the pure acid m. p. 116-117" from 2 g. of p-aminohydratropic acid). Resolution of r- p-Chlorohydratropic Acid.-The solution obtained by boiling 49.5 g. of morphine (I mol.) in 820 C.C. of methyl alcohol was cooled to 42" and 30 g.of r-p-chlorohydratropic acid (1 mol.; prepared from atropic acid McKenzie and Wood Zoc. cit.) were added in one instalment with vigorous stirring. The acid dissolved quickly. When the temperature had fallen to 39' glassy prisms started to separate. After 18 hours in the ice-chest 56 g. of solid were deposited. After the evaporation of the methyl alcohol from the filtrate the acid obtained from the latter gave the following rotation in benzene I = 2 c = 3.924 a? + 3-71' [ c x ] ~ + 47.3", so that the resolution had proceeded markedly after one crystallis-ation. The further progress is indicated as follows 56 g. crystal-lised from 1350 C.C. of methyl alcohol + 35 g. 850 C.C. of alcohol + 19 g. 550 C.C. of alcohol + 13 g. 400 C.C. of alcohol + 9 g.360 C.C. of alcohol + 6 g. pure morphine I-salt. The solution of the crystals in methyl alcohol should be conducted as expeditiously as possible. The progress of the resolution was tested by the polarimetric examination of the acids obtained from the mother-liquors. The morphine salt which separates in rectangular prisms was decom-posed by dilute sulphuric acid the solution extracted with ether nine times and the ethereal extract dried with anhydrous sodium sulphate. The resulting acid (2.1 g.) after drying in a vacuum until constant in weight gave the following rotation in benzene DISPLACEMENT IN THE TROFIC ACID GROUP. PART I. 87 I = 2 c = 3.536 al,fi' - 8-66' [a]:' - 122.4". The acid was crystallised from light petroleum (b. p. 60-80") and gave a value for the rotatory power agreeing with the above within the limits of experimental error Z = 2 c = 3.536 a:" - 8-67' [a]:"' - 122.6." l-p-C~~Zorohydratropic acid m.p. 62-5-63.5" separates from light petroleum in rosettes of glassy plates. It is sparingly soluble in light petroleum and readily so in ether carbon disulphide, benzene and ethyl alcohol (Found C1 = 19.39. C,H,O,Cl requires C1 = 19-240/). I n acetone Z = 2 c = 3.2300 a:.'- - 7-46' [a]::'" - 115.5". Conversion of 1- p-Chlorohydratropic Acid into I-Tropic Acid.-A'-Sodium carbonate (11 c.c.) was gradually added (14 mins.) to a boiling solution of the 1-acid (1 g.) in 20 C.C. of water (reflux). The boiling was continued for 3 hours the cold solution extracted with ether to remove the styrene present' and the tropic acid obtained by acidification with dilute sulphuric acid and extraction with ether.Yield 0.7 g. I n ethyl alcohol Z = 2 c = 2.4348, A similar experiment was carried out with a lzevorotatory acid having [R]:" - 100.7" (1 = 2 c = 3.5504) in benzene. The resulting tropic acid (1.9 g.) gave the following value in ethyl alcohol I = 2 c = 2.716 a;;* - 2-68' [a]:' - 49.3". After one crystallisation of this product from benzene the value for the specific rotation was enhanced in ethyl alcohol 1 = 2 c = 2-548, Conversion of 1- p-Chlorohydratropic Acid into 1-Tropamide.-A solutioii of the I-chloro-acid (0-94 g.) in 60 C.C. of aqueous ammonia (saturated at 0") was kept for a week in a pressure flask at the ordinary temperature. When the solution was transferred to a crpstallising dish needles began to separate.Yield 0.3 g.; in. p. 195-199" (Found N = 8.7. Calc. N = 8.5%). In ethyl alcohol 1 = 2 c = 0.3472 E ~ T - 0.38". I n another experiment the resulting amide gave in ethyl alcohol : The concentrations employed were necessarily very small the aniide being sparingly soluble in water ethyl alcohol acetone, chloroform benzene and ethyl acetate. For the purpose of comparison the amide was prepared from I-tropic acid (obtained by resolution of the 1.-acid with morphine according to McICenzie and Wood) by acting on the ethyl ester with concentrated aqueous ammonia. After crystallisation from acetone the amide had m. p. 195-197.5"; in ethyl alcohol Z = 1, c = 0.377 cxI) - 0.24". Since the polarimetric determinations could be made with very dilute solutions only we have no proof 1 58.1".ai'i 5- - 2.830 1.1 50 -EliS 1 - 3-17" [a]\ - 62.2". I = 2 c = 0.3912 ",F" - 0.43" 88 TROTMAN PREPARATION OF QUATERNARY HYDROCARBONS. that we obtained the optically pure I-tropamide by this method, more especially since there was always the possibility of partial racemisation having taken place during the action of ammonia on the 2-ester. But we think it likely that we obtained the amide nearly pure if not quite so. I-Tropamide would be expected to undergo racemisation in the presence of a small amount of sodium ethoxide (McKenzie and Smith J. 1922 121 1348). When 0.5 C.C. of ethyl-alcoholic potash (0~6518N) was added to a solution giving aD - 0.24" ( I = l), the lzvorotation gradually dropped to zero after 30 hours and r-tropamide (m.p. 170-171") was isolated from the solution. Formation of r-Tropamide from r-Tropic Acid and r- p-Chloro-hydratropic Acid.-r-Tropic acid was converted into its methyl ester from which the amide was prepared by the action of aqueous ammonia. r-Tropamide m. p. 170-171" is sparingly soluble in cold water, ethyl alcohol acetone benzene carbon tetrachloride and light petroleum (b. p. 60-70"). It separates from water in needles (Found N = 8-4. C,H,,O,N requires N = 86%). A solution of r- p-chlorohydratropic acid (4 g.) in concentrated aqueous ammonia (65 c.c.) was kept for 18 days in a pressure flask. On pouring into a dish crystallisation started quickly. The crystals (1.5 g.) were collected and crystallised twice from water, when the pure r-tropamide (m.p. 170-171") was obtained (Found : C = 65.1 ; H = 6-8. The melting point was not depressed when this product was mixed with the r-tropamide obtained from methyl dl-tropate. Calc. C = 65.4; H = 6.7%). One of us (R. C. S.) wishes to thank the Carnegie Trust for the award of a Scholarship which enabled him to take part in. the above investigation. UNIVERSITY COLLEGE D UNDEE . UNIVERSITY OF ST. QNDREWS. [Received November 12th 1924. 82 MCPENZIE AND STRATHERN REACTIONS OF XIV.-Reactions of Displacement in the Tropic Acid Group. Part I. By ALEX. MCKENZIE and ROBERT CAMPBELL STRATHERN. THE proof of the constitution of a-aminohydratropic acid, NH,-CPhMe*CO,H was afforded by Tiemann and Kohler (Ber., 1881 14 1980) by the synthesis from acetophenonecyanohydrin and alcoholic ammonia.McKenzie and Clough (J. 1912 162, 390) in dealing with the Walden inversion described the resolution of the r-acid into its optically active components. The study of the isomeric p-acid NH,*CH,*CHPh*CO,H has not so far been advanced to the corresponding stage. For one reason this acid has no particular interest with relationship to the Walden inversion the amino-group not being directly attached to the asymmetric carbon atom. The identification of the acid has also been attended with some confusion. By the action of aqueous ammonia on p-bromohydratropic acid Merling (AnnaZen, 1881 209 11) obtained crystals m. p. 169-170" (compare also Fittig and Wurster ibid. 1879 195 158).Assuming that the reaction proceeded on the lines CH,Br*CHPh*CO,H + NIP,*CH,*CHPh*CO,H, Merling concluded that he had to deal with p-aminohydratropic acid and the analytical evidence supported this assumption DISPLACEMENT IN THE TROPIC ACID GROUP. PaRT I. 83 Posner however in continuation of his work on the action of hydroxylamine on unsaturated acids obtained from atropic acid an amino-acid which melted and decomposed a t 234" (Ber. 1903, 36 4315; Annulen 1912 389 109). All the evidence pointed to this substance being the true p-aminohydratropic acid. Since Posner (Ber. 1905 38 2316) had shown that the action of ammonia on p-bromo-p-phcnylpropionic acid leads to the formation not of p-amino- p-phenylpropionic acid but of p-hydroxy-p-phenylpro-pionamide he suggested that P-bromohydratropic acid might perhaps exhibit a similar behaviour and that Merling's product was not the amino-acid a t all but was the isomeric tropamide.The latter suggestion has been borne out. In the present paper it is show1 that r-p-chlorohydratropic acid is converted into r-tropamide (m. p. 170-171") by the action of aqueous ammonia. The amide was also prepared directly from ammonia and methyl dl-tropate; it has the same melting point and it possesses the same properties as the compound described by Merling. We have also examined the action of hydroxylamine on atropic acid and find that Posner's p-amino-acid is not the sole product, and that the oxime of phenylacetaldehyde is also formed. It is, however rather curious that the formation of tropic acid by the action of nitrous acid on the amino-acid proceeds in a somewhat irregular manner.The preparation of Z- p-chlorohydratropic acid is described the resolution of the r-acid being effected in methyl-alcoholic solution with morphine as the resolving alkaloid. When hydrolysed the Z-acid which has [.ID - 122.6" in benzene gave a lzvorotatory tropic acid with [ R I D - 58.1" (c = 2.4348) in ethyl-alcoholic solution whereas the value for the optically pure acid is [ R ] ~ - 72.5" (c = 2.578) in the same solvent (NcKenzie and Wood, J. 1919 115 828). Although the chlorine is not directly attached to the asymmetric carbon atom some racemisation had apparently taken place during the substitution of the hydroxy-group for the chlorine.In the light of recent work in a different field (McKenzie and Roger J. 1924 125 214S) the action of ammonia on the E-chloro-acid was of some interest and we have therefore examined it for the purpose of finding if optical activity would persist in the trans-formation into the amide. Experiments showed that a levorotatory tropamide did actually result. Under the conditions employed, the rotation of the product was less than that of the amide obtained by the direct action of ammonia on ethyl I-tropate. This formation of an optically active tropamide has led us t 84 MCKENZIE AND STRATHERN REACTIONS OF conclude that a p-lactone is produced as an intermediate phase, thus : Th Ph Z-H-$-CH,Cl -+ Z-H-$+-CH,Cl + C0,H CO,*NH, Ph Fh Z-H-Y-CH2 -+ Z-H-Y-CH,*OH.cod CO*NH,. Although on those lines the actions do not occur by substitution of groups directly attached to the asymmetric carbon atom some racemisation accompanied the change. This was not altogether unexpected since in tropamide we have the phenyl hydrogen, and carbonyl groups all directly attached to the asymmetric carbon atom a system which a t any rate in the presence of alcoholic alkali is particularly prone to racemisation under certain conditions (McKenzie and Smith J. 1922 121 1348). In the field of optical activity it has been shown by Holmberg (ArEiv Kern. Min. CeoZ. 1917,6 No. 23 pp. 1-33 and earlier papers) that the formation of a p-lactone is a stage in the interconversion of d- and I-malic acids by the Walden inversion. Moreover Z-iodo-succinic acid when treated with ammoniacal silver nitrate gives a mixture of fumaric acid and d-p-malamic acid the latter being formed by the action of ammonia on the d-malo-lactone (compare Walden and Lutz Ber.1897 30 2796; Lutz Ber. 1902 35, 2460 4369; 1908 41 841). Reference has already been made to the fact that Posner obtained r- p-hydroxy- p-phenylpropionamide by the action of ammonia on r- p-bromo- p-phenylpropionic acid. It seems reasonable to suppose that this action proceeds on similar lines to that of ammonia on P-chlorohydratropic acid : ph p Ph Vh H-TBr -+ H T - B r CH,*CO,H CH2*C0,*NH4 CH,*CO CH,*CO*NH,. It will be observed that the p-lactone which is suggested as an intermediate phase is isomeric with the one postulated in the scheme with p-chlorohydratropic acid.Arising from the results recorded in the present paper and more especially from former work by one of us it is desirable that Posner’s observation should be extended to the optically active p-bromo-p-phenylpropionic acids so as to enable the following scheme to be completed: -+ HT-? -+H-$+-O DISPLACEMENT m THE TROPIC ACID GROUP. PART I. 85 dext,ro- dextro-1 zvo - C,H5*CHBr*CH,*C0,H dextro - C GH ,*CH( OH) *CH,*CO*NH, (McKenzie and Humphries J. 1910 97 121; McKenzie and Martin J. 1913 103 112; McKenzie Martin and Rule J., 1014 105 1583; McKenzie and Smith J. 1922 121 1348). Owing to the announcement in the current number of the Pro-ceedings (1924 p. 77) of a paper by Seiiter and Ward entitled " Studies on the Walden iiiversion. Part IX.The influence of the solvent on the sign of the product in the conversion of 3-bromo-p-phenylpropioiiic acids to 3- hydroxy- 9 -phenylpropionamides " TF'C are not meanwhile invcstigatiiig this matter further so far as the action of ammonia on the optically active $-bromo-F-phenyl-propionic acids is concerned. At present we are engaged on a further study of the action of nitrous acid on r-p-aminohydratropic acid and also propose to undertake the resolution of this acid with the object of examining the resulting optically active acids. E X P E R I Jr E 3 T A L. Action of Hydroxyhnine o n Atropic Acid.-Posner (Anncrieih, 1912 389 112) prepared r-$aminohydratropic acid by heating atropic acid with an alcoholic solution of hydroxylsmine for 1 hour. By analogy with his previous work (Ber.190.5 38 2316; compare also McKenzie and Tuclhope J. 1924,125,928) i t might be expected that the formation of the amino-acid would be accompanied by that of phenylncetaldoxiin~.. Kydroxylamine hydrochloride (25 g.) in water (25 c.c.) was adclcd to a solution of sodium (8.5 g.) in ethyl alcohol (280 c.c.), and t hc precipitated sodium chloride removed. Atropic acid (2.3 g.) prepared by the dehydration of atrolactinic acid (McKenzie and Wood Zoc. cit.) was added and the solution boiled for 13 hours. Ammonium carbonate was deposited in the condenser tube. -4fter 3 clays in the ice-chest the amino-acid (8 g.) was collected; it was free from the hydroxylamiiio-acid sirice it did not reduce either Fehling's solution or ainrnoniacal silver nitrate ; i t melted and decomposed a t 222-224" (Posner gives 233").An additional 0.5 g. was obtained by boiling the filtrate for 2& hours longer. After distillation of the bulk of the alcohol an exccss of water precipitated 8 g. of crude phenylaceta!dosime which crystallised from water several times gave the pure substance in lwtrous needles m. p. 103-104" (Found C = 70.9; H = 6.8. Calc., C = 71.1 ; H = 6.7%). Bouvcault and M7ahl (C'ompt. rend. 1002, 134 1147) gave m. p. 103". This osime was isolat cd as follows 86 MCKENZIE AND STRATHERN REACTIONS OF r- P-Aminohydratropic acid is sparingly soluble in water benzene, light petroleum acetone chloroform carbon disulphide and ether. Posner (Ber. 1903 36 4315) described this acid as a-amino-a-phenylpropionic acid (a-aminohydratropic acid) ; its solution in hydrochloric acid was acted on by nitrous acid to give an acid, m.p. 89-90" which was supposed to be atrolactinic acid. Posner was however working with small quantities of material and was misled by the erroneous data of other authors. In a subsequent paper (Annulen Zoc. cit.) the correct constitution of the amino-acid is given. The product which Posner obtained by the action of nitrous acid was apparently impure tropic acid. We have repeated this action several times and find that it proceeds by no means smoothly. The gradual addition of potassium nitrite to the solution of the amino-acid in dilute hydrochloric acid caused the deposition of an oil which was certainly not tropic acid nor did it appear to be a-isotropic acid since it showed no signs of solidifying when nucleated with the latter acid.The aqueous solution was separated from the oil and extracted with ether. By crystallising the resulting acid several times from benzene r-tropic acid was isolated in very small yield (0.1 g. of the pure acid m. p. 116-117" from 2 g. of p-aminohydratropic acid). Resolution of r- p-Chlorohydratropic Acid.-The solution obtained by boiling 49.5 g. of morphine (I mol.) in 820 C.C. of methyl alcohol was cooled to 42" and 30 g. of r-p-chlorohydratropic acid (1 mol.; prepared from atropic acid McKenzie and Wood Zoc. cit.) were added in one instalment with vigorous stirring. The acid dissolved quickly. When the temperature had fallen to 39' glassy prisms started to separate.After 18 hours in the ice-chest 56 g. of solid were deposited. After the evaporation of the methyl alcohol from the filtrate the acid obtained from the latter gave the following rotation in benzene I = 2 c = 3.924 a? + 3-71' [ c x ] ~ + 47.3", so that the resolution had proceeded markedly after one crystallis-ation. The further progress is indicated as follows 56 g. crystal-lised from 1350 C.C. of methyl alcohol + 35 g. 850 C.C. of alcohol + 19 g. 550 C.C. of alcohol + 13 g. 400 C.C. of alcohol + 9 g. 360 C.C. of alcohol + 6 g. pure morphine I-salt. The solution of the crystals in methyl alcohol should be conducted as expeditiously as possible. The progress of the resolution was tested by the polarimetric examination of the acids obtained from the mother-liquors.The morphine salt which separates in rectangular prisms was decom-posed by dilute sulphuric acid the solution extracted with ether nine times and the ethereal extract dried with anhydrous sodium sulphate. The resulting acid (2.1 g.) after drying in a vacuum until constant in weight gave the following rotation in benzene DISPLACEMENT IN THE TROFIC ACID GROUP. PART I. 87 I = 2 c = 3.536 al,fi' - 8-66' [a]:' - 122.4". The acid was crystallised from light petroleum (b. p. 60-80") and gave a value for the rotatory power agreeing with the above within the limits of experimental error Z = 2 c = 3.536 a:" - 8-67' [a]:"' - 122.6." l-p-C~~Zorohydratropic acid m. p. 62-5-63.5" separates from light petroleum in rosettes of glassy plates. It is sparingly soluble in light petroleum and readily so in ether carbon disulphide, benzene and ethyl alcohol (Found C1 = 19.39.C,H,O,Cl requires C1 = 19-240/). I n acetone Z = 2 c = 3.2300 a:.'- - 7-46' [a]::'" - 115.5". Conversion of 1- p-Chlorohydratropic Acid into I-Tropic Acid.-A'-Sodium carbonate (11 c.c.) was gradually added (14 mins.) to a boiling solution of the 1-acid (1 g.) in 20 C.C. of water (reflux). The boiling was continued for 3 hours the cold solution extracted with ether to remove the styrene present' and the tropic acid obtained by acidification with dilute sulphuric acid and extraction with ether. Yield 0.7 g. I n ethyl alcohol Z = 2 c = 2.4348, A similar experiment was carried out with a lzevorotatory acid having [R]:" - 100.7" (1 = 2 c = 3.5504) in benzene.The resulting tropic acid (1.9 g.) gave the following value in ethyl alcohol I = 2 c = 2.716 a;;* - 2-68' [a]:' - 49.3". After one crystallisation of this product from benzene the value for the specific rotation was enhanced in ethyl alcohol 1 = 2 c = 2-548, Conversion of 1- p-Chlorohydratropic Acid into 1-Tropamide.-A solutioii of the I-chloro-acid (0-94 g.) in 60 C.C. of aqueous ammonia (saturated at 0") was kept for a week in a pressure flask at the ordinary temperature. When the solution was transferred to a crpstallising dish needles began to separate. Yield 0.3 g.; in. p. 195-199" (Found N = 8.7. Calc. N = 8.5%). In ethyl alcohol 1 = 2 c = 0.3472 E ~ T - 0.38". I n another experiment the resulting amide gave in ethyl alcohol : The concentrations employed were necessarily very small the aniide being sparingly soluble in water ethyl alcohol acetone, chloroform benzene and ethyl acetate.For the purpose of comparison the amide was prepared from I-tropic acid (obtained by resolution of the 1.-acid with morphine according to McICenzie and Wood) by acting on the ethyl ester with concentrated aqueous ammonia. After crystallisation from acetone the amide had m. p. 195-197.5"; in ethyl alcohol Z = 1, c = 0.377 cxI) - 0.24". Since the polarimetric determinations could be made with very dilute solutions only we have no proof 1 58.1". ai'i 5- - 2.830 1.1 50 -EliS 1 - 3-17" [a]\ - 62.2". I = 2 c = 0.3912 ",F" - 0.43" 88 TROTMAN PREPARATION OF QUATERNARY HYDROCARBONS. that we obtained the optically pure I-tropamide by this method, more especially since there was always the possibility of partial racemisation having taken place during the action of ammonia on the 2-ester.But we think it likely that we obtained the amide nearly pure if not quite so. I-Tropamide would be expected to undergo racemisation in the presence of a small amount of sodium ethoxide (McKenzie and Smith J. 1922 121 1348). When 0.5 C.C. of ethyl-alcoholic potash (0~6518N) was added to a solution giving aD - 0.24" ( I = l), the lzvorotation gradually dropped to zero after 30 hours and r-tropamide (m. p. 170-171") was isolated from the solution. Formation of r-Tropamide from r-Tropic Acid and r- p-Chloro-hydratropic Acid.-r-Tropic acid was converted into its methyl ester from which the amide was prepared by the action of aqueous ammonia. r-Tropamide m. p. 170-171" is sparingly soluble in cold water, ethyl alcohol acetone benzene carbon tetrachloride and light petroleum (b. p. 60-70"). It separates from water in needles (Found N = 8-4. C,H,,O,N requires N = 86%). A solution of r- p-chlorohydratropic acid (4 g.) in concentrated aqueous ammonia (65 c.c.) was kept for 18 days in a pressure flask. On pouring into a dish crystallisation started quickly. The crystals (1.5 g.) were collected and crystallised twice from water, when the pure r-tropamide (m. p. 170-171") was obtained (Found : C = 65.1 ; H = 6-8. The melting point was not depressed when this product was mixed with the r-tropamide obtained from methyl dl-tropate. Calc. C = 65.4; H = 6.7%). One of us (R. C. S.) wishes to thank the Carnegie Trust for the award of a Scholarship which enabled him to take part in. the above investigation. UNIVERSITY COLLEGE D UNDEE . UNIVERSITY OF ST. QNDREWS. [Received November 12th 1924.