Calculated for C10H¿Bг2, 55'94 p. c. Br. Found.. 55.99 55.91 The melting point of this dibromnaphthalene is 63°, but although it approaches my meta-compound in this respect its crystalline formı is perfectly distinct, not only from this, but from all the dibromnaphthalenes hitherto described. It crystallises in beautifully formed oblique rhombic prisms from alcohol, acetone, or petroleum, the latter solvent giving especially large and well-defined crystals. A strong alcoholic solution deposits the substance on cooling as an oil, which does not solidify for some time, so that in order to obtain crystals from this solvent it is better to dilute the alcohol with about 10 per cent. of water, and to crystallise slowly from a weak solution. Constitution of the Dibromnaphthalene melting at 63°.-When one hydrogen-atom in a B-naphthalene-derivative is replaced by Br or NO, the a-position contiguous to the B-substituent appears to be attacked. This has been most clearly proved by the experiments of Jacobson (Ber., 14, 1791), who has shown that when B-acenaphthalide is nitrated an ortho-compound is produced, which by reduction gives a base free from oxygen. The monobrom-3-naphthol obtained by Smith (Chem. Soc. Trans., 35, 789)* has probably a similar constitution. It thus appeared extremely probable, à priori, that the new dibromnaphthalene and the bromnaphthylamine from which it is derived were ortho-compounds. That such is actually the case appears from the following experiments: (1.) Brom-ß-naphthylamine was repeatedly evaporated on a waterbath with dilute nitric acid; a brittle resinous mass finally resulted, from which a large quantity of phthalic acid was obtained. The latter was identified by the melting point of its anhydride and by the fluorescein reaction. Thus the bromine and NH, are in the same benzene-ring. (2.) Brom-3-naphthylamine was dissolved in a small quantity of sulphuric acid diluted with 10 per cent. of water, and the solution was further diluted and well cooled. A diazo-compound was formed by passing nitrous gas through the solution, and the latter was filtered into a large volume of absolute alcohol. After the decomposition of the diazobromnaphthalene and the removal of the excess of alcohol, &c., water caused the separation of an oily body having the boiling point and all the properties of a-bromnaphthalene. Thus the bromine is in an a-position, and the only formulæ possible are: This brom-8-naphthol might furnish orthodibromnaphthalene by the action of PBr, but I have not yet tried this reaction. The first formula is excluded by the fact that the new dibromnaphthalene is isomeric with the one described by me in 1879. The second formula must therefore be adopted, and the present dibromnaphthalene and Cosiner's brom-ß-naphthylamine must be regarded as ortho-compounds belonging to the same series as B-naphthaquinone, nitroso-ßnaphthol, &c. Orthodibromnaphthalene might be expected to occur among the low-melting products of the action of bromine upon naphthalene,* but I have not succeeded in detecting it in the so-called "a-dibromnaphthalene" (m. p. 68-70°) prepared for the purpose of comparison with metadibromnaphthalene. In order to see whether the crystalline form of orthodibromnaphthalene would be disguised by admixture with other isomerides, I made mixtures of this substance both with "a-dibromnaphthalene" (68-70°) and with metadibromnaphthalene, and crystallised from alcoholic solutions. In each case the microscope revealed the rhombic prisms of the ortho-compound amongst the needle-shaped crystals of the other modification. Action of Bromine on Orthobromacenaphthalide. Theory indicates the existence of two homonuclealt tri-derivatives of naphthalene, viz. : * The low melting point (61°) of the a-dibromnaphthalene prepared by Jolia may have been due to the presence of this modification. For the sake of brevity, naphthalene derivatives having the substituents in the same benzene-nucleus may be called homonucleal, and those having the substituents in different benzene-nuclei may be termed heteronucleal derivatives. The simpler expressions, homo-derivatives and hetero-derivatives, do not meet the case, as they II and III may be presumed to be identical, and it became interesting to see whether a second bromine-atom could be introduced into orthobrom-ß-acenaphthalide. If this could bave been effected, the resulting dibromace naphthalide would have furnished on saponification a dibromnaphthylamine which by the elimination of the NH2-group would have given either paradibromnaphthalene or the corresponding meta-compound. These arguments are of course based on the supposition that the second bromine-atom would enter the same nucleus as the other substituents. By a parity of reasoning, such a dibromnaphthylamine on replacing the NH, by Br might be expected to yield one of the tribromnaphthalenes of the above series, i.e., either isomeric or identical with the tribromnaphthalene described by me in the former part of this paper. If identical, the a-position of the second bromine-atom would be demonstrated, and if isomeric, the B-position would be similarly proved. The same considerations obviously apply to the dibromnaphthalene obtainable from this dibromnaphthylamine. By brominating orthobrom-3-acenaphthalide in an acetic acid solution, I have found, however, that the substitution is much more complex than that which gives rise simply to a dibrom-compound. One molecular proportion of the bromacenaphthalide was dissolved in glacial acetic acid, and one molecular proportion of bromine added to the solution. No absorption of bromine had taken place after standing for three days at the ordinary temperature, so the contents of the flask were warmed to about 70-80° for 24 hours, by which time the bromine had disappeared, and a white crystalline substance had separated out. The latter was purified by repeated crystallisation from alcohol, and a compound of constant melting point was at length isolated. This on analysis proved to be a tetrabromacenaphthalide : 0.2939 gram burnt with lime gave 0·4429 gram AgBr. Calculated for C10H3Bг.NH.C2H2O, 63.87 p. c. Br. Found, 64-10. The alcoholic mother-liquors appear to contain a complex mixture of bromine-derivatives from which nothing definite has as yet been isolated. The tetrabrom-compound forms very minute white needles melting at 138°, It is freely soluble in warm and but slightly soluble in cold alcohol; it dissolves in cold ether, acetone, chloroform, and benzene, and very freely in carbon disulphide. An experiment was made in order to obtain the corresponding tetrabromnaphthylamine, but three days' boiling with a syrupy solution of caustic potash failed to remove the C2H3O-group. would indicate only dissimilarity of the substituents. It seems to be a very general Jaw in the naphthalene series that homonucleal di-derivatives have lower melting points than their heteronucleal isomerides. By brominating paranitracenaphthalide, I have obtained a bromnitracenaphthalide crystallising in pale yellow needles, melting at 224 -225°, and isomeric with the body obtained by Liebermann and Scheiding (Ber., 8, 1108) by nitrating parabromacenaphthalide, which melts at 232°. Experiments with a view to determine the constitution of this and of other nitrobrom-derivatives are in progress. These researches have been conducted in the laboratory of the Atlas Works at Hackney Wick, and it gives me great pleasure to be able once again to record my obligations to Messrs. Brooke, Simpson, and Spiller. II.-On the Constitution of Lophine. (Second Notice.) By FRANCIS R. JAPP, M.A., Ph.D., Assistant Professor of Chemistry in the Normal School of Science, South Kensington. IN No. 11 of the Berichte, 1882, p. 1493 (see also this Journal, 1882, Abstracts, p. 1063) Radziszewski communicates a new synthesis of lophine by the interaction of benzil, benzaldehyde, and ammonia. This reaction corresponds with the synthesis of para-hydroxylophine from benzil, parahydroxybenzaldehyde, and ammonia, described by Japp and Robinson (Ber., 1882, p. 1268; this Journal, 1882, p. 326). In discussing his synthesis, Radziszewski comes to the conclusion that lophine has the formula C&H-C-N CH-CH, CH-C-N and rejects the formula C&H-C-NH C-CH, C&H-C-N proposed by Mr. Robinson and myself. In the latter formula, lophine is represented as belonging to the class of the anhydro-bases described by Hübner. I shall, therefore, in the present paper, refer to this formula as the "anhydro-base formula " of lophine. This formula was based chiefly upon certain analogies drawn from the reactions of phenanthraquinone with aldehydes and ammonia. I will now endeavour to show that these analogies are well founded, inasmuch as benzil really yields with aldehydes and ammonia the same classes of compounds as phenanthraquinone, though not always under the same conditions; further, that the anhydro-base formula explains the known reactions of lophine more consistently than that of Radziszewski; and lastly, I shall describe an experiment which, though not absolutely conclusive, affords a strong presumption in favour of the anhydro-base formula. In order that what follows may be more readily understood, I will re-state here the two fundamental reactions of phenanthraquinone with aldehydes and ammonia : CH,−CO CH-C-O I. + R-CHO + NH, = | C-R'+ 20H2. | | = C.H.-CO C&H1-CO CH-C-N C.H.-C-NH C-R'+30H2. II. | | +R–CHO+2NH= || CH-C-N The first reaction occurs with non-hydroxylated, the second with hydroxylated aldehydes of the benzene series. This explains sufficiently why we resorted to the indirect method of employing a hydroxyaldehyde, and thus preparing first a hydroxylophine, instead of, like Radziszewski, acting upon benzil with benzaldehyde in presence of ammonia, and thus preparing lophine directly; for, following the analogy of the phenanthraquinone reactions, we must have expected in the latter case to obtain the compound CH-C-O. C-CH, C&H-C-N by a reaction corresponding with that expressed in equation I.* In fact, previously to the publication of Radziszewski's paper, we had tried the reaction which he describes. We hoped to obtain the above oxygenated compound, and were surprised to find that the reaction. yielded lophine. We conducted the experiment under conditions somewhat different from those adhered to by Radziszewski: instead of saturating an alcoholic solution of benzil and benzaldehyde with ammonia at a temperature of 40-50°, we heated molecular proportions of benzil and benzaldehyde with aqueous ammonia under pressure-on one occasion at 100°, on the second at 150°. In the first * We were the more justified in this expectation, inasmuch as chrysoquinonealso a double ketone-yields, with benzaldehyde and ammonia, benzenylamidochrysole. |