and to this compound the name ethylic a-phenanthroxylene-isocrotonate might be given. That it is not the oxygen-atom of the phenanthroxylene-group which is removed during the reduction, is very clearly shown by the behaviour of the reduction-product towards caustic alkalis (vide infra). The same reduction-compound is obtained when the condensationproduct is treated in acetic acid solution with zinc-dust; but the yield by this method does not appear to be so good as when hydriodic acid is employed. The reduction-compound forms with bromine in acetic acid solution a compound-probably additive-which has not been examined. On oxidation with a chromic mixture the reduction-compound yields phenanthraquinone. On heating between watch-glasses it yields a sublimate of a new compound, in the form of white needles, fusing at 213°, the examination of which is described later. A portion of the substance remains as a charred mass. By heating ethylic phenanthroxylene-aceto-acetate with fuming hydriodic acid and amorphous phosphorus to 200°, a second reductioncompound was obtained, in the shape of an acid. Mineral acids precipitated it from the solutions of its salts as an amorphous substance, insoluble in all the ordinary organic solvents. Neither the acid nor its salts could be obtained in a crystallised condition, so the further investigation of this substance was abandoned. Behaviour of Ethylic Phenanthroxylene-isocrotonate towards Caustic Alkalis.-Solutions of alkaline carbonates are without action upon this compound, but dilute caustic potash dissolves it readily on gently warming. On adding hydrochloric acid to an alkaline solution thus prepared, a new organic acid was precipitated. This acid was almost insoluble in alcohol and other organic solvents of low boiling point, but boiling phenol dissolved it readily, and on carefully diluting the hot solution with alcohol, the acid separated in fine colourless needles, which, after washing with alcohol and drying, fused at 295°. This fusing point was not altered by a second crystallisation from phenol. The results of analysis agreed with the formula C18H1O4:— The caustic alkali had therefore not merely saponified the ethylic phenanthroxylene isocrotonate, but had at the same time effected the addition of the elements of a molecule of water to the acid thus produced: C20H1603 + OH2 = C18H12O3 + C2H2O...... (Saponification), and CH2O + OH = C,H,O... New acid. .......... (Addition of water). An examination of the salts of this acid disclosed the remarkable fact that the acid was dibasic, assuming, as is unavoidable from the mode of formation of the compound and from its reactions, that the molecular formula is that given above, and not half this formula. Silver Salt. This was obtained as a white crystalline insoluble precipitate by adding silver nitrate to a neutral solution of the ammonium salt. It yielded the following numbers on combustion : Barium Salt.-This salt was prepared by precipitating the ammonium salt with barium chloride. It formed a white, insoluble, crystalline powder. Determinations of barium and water of crystallisation gave numbers agreeing with the formula C1H12O,Ba,20H,-Ba (in crystallised salt): Found, 29-39; calculated, 29:46 per cent. OH, (not entirely expelled below 220°, at which temperature the salt becomes slightly dark-coloured): Found, 8:33; calculated, 7.74 per cent. slight decomposition had therefore probably taken place. At 180° only 1 mol. of water of crystallisation is expelled. A The acid may therefore be formulated C16H12(COOH)2. It does not, University of L like the two compounds already described, yield phenanthraquinone on oxidation. We repeated the oxidation under varying conditions, but on no occasion was a trace of the easily recognisable quinone formed. This negative result is of importance, as it renders it probable that in the formation of the dibasic acid the closed lateral chain of the phenanthrene-group is severed. An inspection of the formula of ethylic phenanthroxylene-isocrotonate shows that it is in fact not possible to obtain from this substance by saponification and simultaneous addition of the elements of water a dibasic acid without thus severing the closed chain. The quantity of substance at our disposal was insufficient to allow of our studying the oxidation-products of the acid. Heated between watch-glasses, the acid yielded a sublimate of colourless needles fusing at 213°, identical with the sublimate obtained from ethylic phenanthroxylene-isocrotonate. A portion of the substance was charred in the process; but the yield of sublimate was better than in the former case. A quantity of this sublimate was therefore prepared from the acid. It was purified by crystallisation from boiling alcohol, in which it dissolves readily, separating out almost entirely on cooling. It was thus obtained in fine colourless silky needles, fusing, as above, at 213°. Analysis gave results agreeing with the formula C1H10O: We were not able to examine this substance further, except to ascertain that it is insoluble both in caustic alkalis and in hydrogen sodium sulphite. The small quantity at our disposal sufficed only for a single analysis, and the preparation of the substance in any quantity would be a work of very great labour. To judge from the ease with which the substance sublimes, the above simple formula is probably also its molecular formula. Constitution of the Dibasic Acid.-There are two ways in which a compound of the formula of ethylic phenanthroxylene-isocrotonate may by the action of caustic alkali be converted into a dibasic acid. The simplest explanation of the phenomenon consists in supposing that, along with the saponification of the COOCH, group, the carbonyl-group of phenanthroxylene is separated from the other carbon-atom of the closed lateral chain; potassoxyl then attaches itself to the vacant affinity of the carbonyl, converting it into COOK, whilst the remaining hydrogen-atom from the potassium hydroxide satisfies the vacant affinity of the other carbon-atom of the severed lateral chain. The acid would thus possess the constitution represented by the formula The second mode of regarding the reaction is to suppose that an intramolecular migration takes place similar to that in which phenanthraquinone is converted by the action of caustic potash into diphenyleneglycolic acid, a reaction in which the phenanthraquinone takes up the elements of a molecule of water, whilst one of its carbonyl-groups is converted into carboxyl. On this assumption the dibasic acid would be a derivative of diphenylene-methane. We think, however, that this last view is to be rejected. The fact that both ethylic phenanthroxylene-isocrotonate and the dibasic acid yield under the influence of heat the same compound, C1H10O, points to an analogous constitution of the two compounds, a condition which is not fulfilled by representing the one as a derivative of phenanthrene and the other as a derivative of diphenylenemethane. Rejecting therefore this view, there remains for our acceptance the formula above given. A possible constitution for a compound C1HO, derived both from ethylic phenanthroxylene-isocrotonate and from the dibasic acid, would be CH-CH2 Such a compound would stand in the same relation to phenanthraquinone in which deoxybenzoïn stands to benzil. An investigation of the reactions of this compound would doubtless have been desirable, and would possibly have thrown light upon the constitution of the compounds from which it is derived. We have already alluded to the difficulties which would attend such an investigation. The chief value of the facts here described lies in their bearing upon the constitution of phenanthraquinone. The formation of such a compound as ethylic phenanthroxylene-aceto-acetate from ethylic aceto-acetate and phenanthraquinone, furnishes a very strong argument in favour of Fittig's formula for phenanthraquinone as against that of Graebe. The investigation will be continued. 35 V.-Note on the Preparation of Diphenylene Ketone Oxide. By W. H. PERKIN, Ph.D., F.R.S. WHILST making experiments with the hope of preparing the anhydride CH of salicylic acid, | >O, a quantity of salicylic acid was heated with CO acetic anhydride. On boiling the mixture, the acid dissolved, acetic acid and the excess of acetic anhydride used distilling off. A viscid liquid then remained in the retort, solidifying on cooling to a glass-like mass, undoubtedly consisting chiefly of salicylide. On submitting this to distillation, a considerable quantity of an oily product came over, solidifying in the neck of the retort to a crystalline mass. This on being washed with alcohol, and purified two or three times by crystallisation from that solvent, was obtained in the form of fine pale-yellow needles. On analysis this substance gave the following numbers: I. 0-1450 gram of substance gave 0·4222 of CO2 and 0.0571 of OH. II. 0.1488 0.4335 0.0559 These give percentages agreeing with the formula C13H2O2. This substance is evidently the same as that obtained by Merz and Weith (Ber., 14, 187) by the oxidation of methylene-diphenyl oxide, also by R. Richter (J. pr. Chem., N.F., 23, 349) by distilling basic potassium salicylate with phosphorus oxychloride. This latter process, however, yields it, so far as I have experimented with it, in only comparatively small quantities, whereas by the method above described, from 30 to 40 per cent. of the theoretical quantity is obtained. Its formation from salicylide may be represented thus: Co This substance being now obtainable with comparative ease by the above process, my son, Mr. A. G. Perkin, has commenced the study of its derivatives, and of the secondary bodies which are also obtained in its preparation from salicylic acid. VOL. XLIII. D |