THE RELATIONSHIP OF CORYDALINE TO BERBERINE. 157 XV.-TJle Relationship of Corydaline to Berberine. Bei*beridic Acid. By JANES J. DOBBIE, M.A., D.Sc., and ALEXANDER LAUDER, B.Sc. PERKIN (Trans., 1890, 57, 992) has proposed the following alternative formulae for berberine, expressing the opinion that (I) is the more probable of the two : L I. 11. In the preceding paper, we have shown that the constitution of corydaline can be represented by a formula similar to I, and assum- ing the correctness of this formula for corydaline, that the absence of compounds corresponding to berberal, C,,HI7O7N, berberilic acid, C,,H,,O,N, &c., from amongst the decomposition products of cory- daline is explained.* Whilst the absence of certain decomposition products is satisfactorily accounted for, the similarity of the formulae assigned to the two alkaloids suggested the possibility of obtaining from berberine an acid corresponding to corydic acid, and, as a matter of fact, we found no difficulty in preparing the expected acid by a method similar to that used in the preparation of corydic acid. For convenience of reference, we shall provisionally term the substance so obtained berberidic acid.Ten grams of berberine nitrate were suspended in two Iitres of dilute nitric acid (1 in 20) and heated a t the temperature of the water-bath until completely dissolved. When the solution cooled, a small quantity of the new acid was deposited as a yellow, crystalline precipitate. This was filtered off, the solution neutralised with ammonia, concentrated, and precipitated with silver nitrate.The silver precipitate was decom- posed with sulphuretted hydrogen and the acid separated by fractional crystallisation from a more soluble substance not yet examined, which was formed along with it. I n crystallising the acid, a considerable amount of t a r q matter separated out. The acid was finally freed from this and obtained in a pure state by dissolving in sodium hydr- oxide and precipitating with hydrochloric acid. I n later preparations, * For further comparison of berberin with corydaline, see preceding paper,158 DOBBIE AND LAUDER: THE RELATIONSHIP OF the purification was greatly facilitated by fractional precipitation with silver nitrate, the first fraction carrying down most of the tar. The subsequent fractions were light in colour and practically pure.The yield of purified acid amounted to about 20 per cent. of the berberine nitrate used. Berberidic acid crystallises from water in radiating tufts of yellowish-brown, prismatic crystals, which have a pure yellow colour when powdered. It contains no water of crystallisation. When heated in a capillary tabe, it darkens at about 235' and remains without further change, so far as can be seen, until 285', when it melts with decomposition. It was dried at 100' and analysed, with the following results : 0,2637 gave 0,5925 CO, and 0-0866 H,O. C=61*28; H=3.61. 0.3243 ,, 13.0 C.C. nitrogen at 16' and 761.5 mm. N=4.75. 0.2808 ,, 11.0 C.C. ,, 13' ,, 751 mm. N=4.63. Cl,H,,O,N requires C = 61-34 ; H= 3-51 ; N = 4.47 per cent. 0.2831 ,, 0.6344 CO, ,, 0.0917 H,O.C=61*12 j H=3.59. Berberidic acid is insoluble in cold and only sparingly soluble in boiling water. It is very sparingly soluble in boiling alcohol and in- soluble in ether or chloroform. It dissolves easily in sodium hydroxide to a dark blood-red solution, from which it is precipitated by hydro- chloric acid, All its salts, with the exception of the two silver salts, appear to be soluble. The normal silver salt is obtained by precipitating a solution of the acid, which has been neutralised with ammonia, with silver nitrate. A curdy, yellow pre- cipitate is obtained, which darkens on exposure to light. This salt was repeatedly prepared and analysed without exact results being obtained, owing, apparently, to admixture with the acid salt.The acid silver salt is prepared by precipitating an aqueous solution of the acid with silver nitrate. The curdy precipitate so obtained is filtered, washed, and purified by repeated recrystallisation from water. It is finally obtained in stellate clusters of beautiful, yellowish-brown needles. On heating, it decomposes suddenly with evolution of thick, brown vapours. After being dried at loo', it was analysed with the following results : Berberidic acid is dibasic. 0.2470 gave 0.0828 AgCl. Ag = 25 *23. 0,2616 ,, 0.0655 AgCl. Ag = 25.04. C,,H,,O,NAg requires Ag = 25.71 per cent. When berberidic acid is heated with concentrated hydrogen iodide solution, no methyl iodide is evolved, a fact which proves that in the formation of this acid the ring of the berberine molecule containing the methoxyl groups is destroyed.CORYDA1,INE TO BERBER.INE, BERBERIDIC ACID.159 Oxidation of Berberidic Acid with Potassium Permanganata-Five grams of berberidjc acid were boiled with a dilute solution of per- manganate until the permanganate was no longer reduced. The solution was filtered from the manganese oxide, concentrated, and pre- cipi tated with silver nitrate. The silver precipitate was decomposed with sulphuretted hydrogen and the filtrate from the silver sulphide evaporated to dryness. The residue was repeatedly exhausted with hot absolute alcohol, in which a considerable part of it dissolved. The portion of the residue insoluble in hot alcohol dissolved readily in boiling water, from which it separated on cooling in prismatic crystals.The acid so obtained was decolorised by boiling with charcoal and purified by repeated recrystallisation from water. I t melted at 235' or 242', according to the rate of heating. It dissolved with difficulty in cold, but was readily soluble in boiling, water ; it was insolubb in ether or chloroform. Its aqueous solution gave an orange-red colora- tion with ferrous sulphate. The acid agrees in every particular with C0,H /\CO,H \N berberonic acid, co HI ) , which was obtained by Weidel (Ber., 1879, 12, 410) by the direct oxidation of berberine with concentrated nitric acid. The melting point of berberonic acid is variously given a t The normal silver salt, which is almost insoluble in water, was pre- pared by precipitating a solution of the acid, previously neutralised with ammonia, with silver nitrate.After bemg dried at looo, it was analysed with the following result : 238--.242'. 0.2978 gave 0.1798 Ag. Ag=60.38. C,H20,NAg, requires Ag = 60.88 per cent. The presence of hydrastic acid amongst the decomposition products of berberidic acid has not yet been proved. By dissolving berberidic acid in potassium carbonate and oxidising it with potassium perman- ganate a t the ordinary temperature, a small quantity of a substance was obtained as a scum on the surface of the strongly alkaIine solu- tion. From its insolubility in potash, we suspected that this sub- stance might be w-aminoethylpiperonylcarboxylic anhydride, which is insoluble in alkaline solutions. On examination, we found that it agreed in every particular with the anhydride in its neutral reaction, solubility, peculiar mode of crystallisation, and behaviour with mercuric chloride. As the amount of substance obtained was too small to admit of complete purification, the melting point observed was slightly lower than that given by Perkin. Berberidic acid clearly bears the same relation to berberine that;160 FORSTER AND MICKLETHWAIT : STUDIES IN THE corydic acid bears to dehydrocorydaline. Since it contains no methoxyl groups, it follows that the ring of the berberine molecule which yields hemipinic acid is destroyed in its formation. The occurrence of o-aminoethylpiperonylcarboxylic anhydride and berberonic acid amongst its oxidation products proves that it contains the three remaining rings and that its constitution may therefore be expressed by the formula : UO,H By oxidising berberidic acid with potassium permanganate at the ordiiiary temperature, a yellow derivative is obtained like that obtained from corydic acid by similar treatment.As berberine, unlike corydaline, can be obtained at comparatively low cost, we have undertaken a more thorough investigation of ber- beridic acid, which we hope will throw further light on the constitu- tion of both alkaloids, and especially on the relation between the constitution and colour of some of their derivatives. UNIVERSITY COLLEGE OF NORI-H WALES, BANGOR.