the formula CH,,NO,. H2O. We believe that much of the difficulty encountered in the determination of the composition of this base rests in the fact that purification is effected with difficulty. Having prepared a considerable quantity of chelidonine from Stylophorum, we took up its study mainly for the purpose of verifying the very latest results upon its composition. For the determination of water of crystallization a small quantity of the freshly crystallized alkaloid was heated in an air-bath. at a temperature of 100° to constant weight. Henschke used heat of 125° C., but at this temperature our product became gradually darker. Since neither Henschke nor Selle make mention of this in their report we were of the impression that our product must still be contaminated with impurities, therefore, the colorless alkaloid was recrystallized twice from alcohol and again heated but with the same result. Then a lower heat was tried and sufficiently prolonged to cause an appreciable loss of water, but again with the same change in color. Fearing that this coloration indicated a slight decomposition and therefore would vitiate results, an attempt was made to remove the water in a vacuum desiccator over sulphuric acid and phosphoric anhydride until the substance lost no weight. A vacuum was maintained (with interruptions for weighing) for fully a month and it was noticed as increasing quantities of water were removed the alkaloid became more yellow. Unfortunately we could not complete the experiment because of an accident but it was plainly evident that with removal of water of crystallization even in the cold, discoloration takes place. Since light seems to hasten the change, there is in progress at this time an experiment to determine whether all the water is removed and also whether discoloration takes place in the dark. A sample heated to constant weight at 100° for about fourteen hours lost 4.5 per cent. of its weight. According to the formula. CHNO,.HO, the theoretical amount of water is 4.8 per cent. Several combustions were made with the air-dried material as well as with that dried to constant weight but the results were not as concordant as was deşired and were therefore rejected. Excellent results were obtained with the nitrate which is easily made pure and which contains no water of crystallization. In the estimation of the acid in the hydrochloride by means of silver nitrate the filtrate from the precipitated silver chloride separated large needles of the nitrate of chelidonine. After recrystallizing several times from hot water and drying in a desiccator, the following results of the combustions were obtained : I. 0.333 gram chelidonine nitrate gave 0.7044 gram carbon dioxide and 0.1524 gram water. II. 0.350 gram chelidonine nitrate gave 0.1532 gram water. III. 0.317 gram chelidonine nitrate gave 0.6738 gram carbon dioxide and 0.1354 gram water. Carbon.. Hydrogen. Calculated for Found. Nitrogen was determined by the Dumas as well as by the Kjeldahl method. By the former method the following results were obtained : I. 0.4926 gram of anhydrous chelidonine gave 0.0211 gram of nitrogen, or 4.27 per cent. II. 0.551 gram of anhydrous chelidonine gave 0.022 gram of nitrogen, or 3.97 per cent. Theory requires 3.96 per cent. By the latter method the following results were obtained: I. 0.412 gram of chelidonine gave 3.33 per cent. of nitrogen. II. 0.486 gram of chelidonine gave 3.40 per cent. of nitrogen. Theory requires 3.77 per cent. The hydrochloride was prepared and purified as stated above. Air-dried material heated in an oven for several hours did not lose weight; therefore there is no water of crystallization. The acid was determined in the customary way with silver nitrate and weighing the washed and dried silver chloride. I. 0.455 gram of chelidonine hydrochloride gave 0.1656 gram silver chloride. II. 0.3406 gram of chelidonine hydrochloride gave 0.1264 gram silver chloride. The analyses of the gold and platinum double salts verified the formula obtained from combustions of the nitrate. The gold salt was readily made by precipitating a hot acid solution of the hydrochloride of chelidonine with an excess of a 2 per cent. solution of gold chloride. The double salt separated as a voluminous, orange-red precipitate which was collected on a filter, thoroughly washed and dissolved in hot alcohol. Beautiful violet-red crystals bunched in rosettes separated in a short time and these were re crystallized and dried in a desiccator. Heating in an oven at 100° for several hours did not cause it to lose weight. I. 0.0774 gram of the gold salt yielded, upon incineration, 0.022 gram gold. Gold...... Calculated for .... 28.1 28.42 The platinum salt was made in the same manner, using a 5 per cent. solution of platinum chloride. The double salt separated as a yellowish precipitate which became denser and deeper colored when reprecipitated for purification. We did not succeed in obtaining crystals of this compound from ethyl or methyl alcohol. One small fraction that had been standing in methyl alcohol for several days had changed to fine silky needles but we were not able to repeat the operation. Heating to 100° to constant weight removes 3.4 per cent. water. I. 0.136 gram of the water-free double salt left, upon incineration, 0.0236 gram platinum. Platinum Calculated for Water.... 3.4 Action of Ethyl Iodide.-Five grams of pure dried material were heated in a bomb tube with an excess of pure ethyl iodide at 130-140° for four hours. When cold, the tube was opened and only a slight pressure noticed. In the bottom of the tube there was an insoluble portion yellowish in color and above it a transparent, light red liquid. The excess of ethyl iodide was driven. off and the residue dissolved in boiling alcohol. It was filtered and set aside in a small Erlenmeyer flask. While crystals began to appear after a considerable length of time we found that layering with an equal volume of ether as employed by Henschke hastened crystallization materially. Tufts of fine, silky needles formed on the side of the flask and at the contact zone of the two liquids. Successive crops were obtained and all the combined. fractions recrystallized from the same solvents several times but with loss of considerable material. Enough was left, however, to determine the iodine content with silver nitrate. The new compound was found to be free of water of crystallization. I. 0.2038 gram of the compound gave 0.0948 gram silver iodide. Iodine Calculated for 24.94 Found. Long contact of potassium hydroxide with a solution of this compound had no effect since the same compound with identical melting-point was again obtained in almost quantitative amount. This alkaloid is probably a tertiary base, as was shown by Henschke. For the determination of methoxyl groups a modification of Zeissel's well-known method was employed, using ground-glass joints throughout the entire apparatus. After carrying on the operation for fully an hour not the least turbidity was noticed in the silver nitrate solution. Methoxyl groups are therefore absent. By pouring a solution of the sulphate of chelidonine into a large excess of Wagner's reagent, an abundant chocolate-colored precipitate formed which, upon shaking vigorously, became denser and settled to the bottom. The collected precipitate was thoroughly washed and then dissolved in hot methyl alcohol and set aside in flat crystallization dishes. Two distinct forms separated, one in light red needles and the other in almost black prisms. These were separated mechanically and each recrystallized several times. from hot methyl alcohol. When dried in a desiccator and the iodine estimated with thiosulphate solution, the light red crystals gave figures that agree with the formula for the triiodide CHNO,.HI.I,. The black prisms appear to have the formula CH,,NO.HI.I. Finally a solution of the free alkaloid chelidonine, dissolved in absolute alcohol, gave an optical rotation of [a] = + 115°24'. ALKALOID II. This one is second in abundance and crystallizes in distinct needles as shown in Fig. 2. The needles which Selle obtained in such small quantity that only the melting point (193°-195°)' could be made are probably the same substance. Since we have not found in literature anywhere a description of an alkaloid possessing the properties of this one we have decided to designate it by the name stylopine. All the fractions of the needles melting in the neighborhood of 200° were united and purified in the same manner employed with chelidonine. A pure product was obtained that possessed the constant melting-point of 202° (uncorr.). This alkaloid is almost insoluble in hydrochloric acid, forming fine needles of the salt when strong hydrochloric acid is added to a 1 Arch. d. Pharm., 228, 108. solution of the acetate. It is also insoluble in dilute sulphuric acid which serves excellently as a means of separation from chelidonine. The free alkaloid is very soluble in glacial acetic acid, much less so in dilute acid. The nitrate separates from aqueous solutions in very small clusters of needles which in mass appear almost jellylike, or gelatinous. Precipitates were obtained with well-known reagents as follows: bustions of the pure dry substance gave the following results : I. 0.1926 gram stylopine gave 0.4712 gram carbon dioxide and 0.0958 gram water. II. 0.1718 gram stylopine gave 0.0864 gram water. III. 0.1704 gram stylopine gave 0.4206 gram carbon dioxide and 0.0854 gram water. IV. 0.2714 gram stylopine gave 0.6680 gram carbon dioxide. Four nitrogen determinations gave the following results : I. 0.1768 gram stylopine gave 0.007478 gram nitrogen. II. 0.177 gram stylopine gave 0.007292 gram nitrogen. III. 0.1824 gram stylopine gave 0.00698 gram nitrogen. IV. 0.1316 gram stylopine gave 0.005578 gram nitrogen. The hydrochloride was made in the usual manner and purified by recrystallization from hot water. It contains no water of crystallization. For the determination of the formula of this salt the acid was estimated by precipitating with silver nitrate and weighing the washed and dried silver chloride. Only one estimation was made. |