280 HOMER AND PURVIS: THE ABSORPTION SPECTRA OFXX1X.-The A bsorption Spectra of Naphthalene andof Tetrame t h y lnaph t ha1 ene.BY ANNIE HOMER and JOHN EDWARD PURVIS.IN a paper on the absorption spectra of the hydrocarbons isolatedfrom the products of the action of aluminium chloride on naphthalene(Homer and Purvis, Trans., 1908, 93, 1319), we described theabsorption spectrum of a substance having the empirical formulaC14H,,. From chemical con5iderations it had been suggested that* Air-dried. t Dried at 100NAPHTHALENE AND OF TETRAMETHYLNAPHTHALENE. 281this hydrocarbon was an alkyl, probably a tetramethyl, derivative ofnaphthalene, and it was thought that a comparison of the absorptionspectrum of this hydrocarbon with that of naphthalene would furnishadditional evidence as to its constitution.Hartley, in his earlier work (Trans., 1881, 29, 1531, showed thatthe absorption spectrum oE naphthalene was characterised by fourbands in the ultraviolet region of the spectrum.In a later paper,Hartley (Trans,, 1886, 47, 685) showed that there were four bands,the positions of which differed from those of the earlier observations.The method employed by Hartley differed from that used by us, anddescribed in the previous paper (Zoc. cit.). For purposes of comparisonit was necessary for us to study the absorption spectrum of naphths-lene under the same experimental conditions as those employed in thestudy of the hydrocarbons under investigation. Our results showedthat for N/LOOO-solutions in alcohol, the absorption curve fornaphthalene, as observed by Hartley in his first paper, showed fourcharacteristic bands.The mean oscillation frequencies of thesebands did not coincide with those given by Hartley, but we did notcomment on the discrepancies at the time of our previous work.The curve for N/lOOO-solutions of the hydrocarbon Cl4H1, alsoshowed four bands, similar in character to the naphthalene bands,although less persistent and further shifted towards the red end ofthe spectrum (Fig. 1).The mean oscillation frequencies of the bands were :For naphthalene, Hartley, first paper ...... 3508 3690 3840 392139 second paper ... 3211 3273 3369 3849previously published).. 3500 3620 3765 3900For CI4Hl6, Homer and Purvis ............... 3434 3562 3680 3776From these results we concluded that the substance C14H16 wasprobably tetramethylnaphthalene, as had been previously suggested(Homer, Trans., 1907, 91, 1103).Since the publication of these results, Baly and Tuck (Trans.,1908, 93, 1902), in their work on the absorption spectra of aromatichydrocarbons,jhave shown that for naphthalene thereare three character-istic bands having mean oscillation frequencies : 1/h = 3125, 1/X = 3220,and 1/X=3700.The bands 1/X 3125 and I/h 3220 only appearin strong solutions or in proportionately greater thicknesses of dilutesolutions. The band 1/A 3700 is a broad band, and corresponds withthe four bands observed by Hartley in his first paper (Trans., 1881,39, 153) and also by us.After reading Baly and Tuck’s paper, it seemed necessary for us tore-examine the absorption spectrum of the hydrocarbon C,,H1,.Forif, as we had previously surmised, the hydrocarbon C14H16 is a tetra-BS Homer and Purvis (no282 HOMER AND PURVIS: THE ABSORPTION SPECTRA OFmethyl derivative of naphtbalene, there should be indications of bandscorresponding with the naphthalene bands, 1/A 3125 and l / A 3230,observed by Baly and Tuck. The absence of such bands would, accord-ing to these observers, point to a difference between the constitutionof naphthalene and of the hydrocarbon C,,H,,.At the time of our previous experiments we had only suf3ticient ofthe hydrocarbon with which to make a N/lOOO-solution. This strengthwas compared with a similar N/lOOO-solution of naphthalene, Thesesolutions gave the four bands already described, and correspondingwith the broad band, 1 / A 3700, of Baly and Tuck.Since then theFIG. 1.Oscillation frequencies.33 34 35 36 37 38 39 40 41 42Full curve : N/lOOO-solution of naphthalene in alcohol.Dash ,, : N/lOOO-solution of CI4H,, in alcohol.hydrocarbon has been made in larger quantity, and we have been ableto compare its iVj’10-solution with a similar solution of naphthalene.The results have been plotted in the accompanying curves (Fig. 2).It will be seen from a glance a t the curves, that for N/lO-solutionof naphthalene there are two bands similar to those obtained byBaly and Tuck. The curve for the hydrocarbon C14H16 also showstwo similar bands, less persistent and further shifted towards thered end of the spectrum than the naphthalene bands.The mean oscillation frequencies of the bands are :Naphthalene, Baly and Tuck ........., .... ... ... ... .. . 3125 3220), N/10-solution, Homer and Purvis ... 3122 3225C14Hlsr N/lO-solution, $ 9 3 ) 9 9 * * I 2638 277NAPHTHALENE AND OF TETRAMETHYLNAPHTHALENE. 283The results of our investigation may be summarised as follows :I. The absorption curves for both N/lOOO- and N/lO-solutions ofthe hydrocarbon CI,HI6 are similar to the curves for naphthaleneThese results are in accordance with our previous suggestion thatthe substance is a tetramethyl derivative of naphthalene.11. Our experiments confirm the observation of Bnly and TuckFIG. 2.Oscillation frequencies.26 28 30 32 34 36Fnll curve : N/IO-solution 04 naphthalene i n clkohol.Dash ,, : N/lO-soZution of C14H,, in alcohol.with regard to the preqence of bands corresponding with l / X 3125and l / h 3220 in naphthalene and its derivative.111. Hartley in his first paper represented the naphthalene curvefor dilute solutions as being characterised by the presence of fournarrow bands. Baly and Tuck from their results consider that thereis only one broad band, 1/X 3700, which corresponds with Hartley’284 BARGER AND EWINS:four bands. The curves obtained by us for N/lOOO-solutions ofnaphthalene, and of its derivative, CI4Hl6, shows four narrowbands. These observations wpport Hartley’s first interpretation ofthe absorption curve of naphthalene.UNIVERSITY CHEMICAL LABORATOKY,CAMBRIDGE