598 



NATURE 



{April 1 8, 1889 



■their fundamental constituents— the atoms. In the same way it 

 was possible that the results obtained by Raoult's method by 

 means of observations on the behaviour of molecular complexes 

 might afford the means of deducing the relative magnitudes of the 

 fanda nental molecules comprising the complexes:, but not of the 

 actual complexe-; operated with. Mr. Crompton drew attention 

 to Heckmann's recent experiments on the lowering of the freezing- 

 point ; these fhov that the true molecular weight was only 

 obtained when solutions were used the concentration of which 

 was allowed to vary only within certain narrow limits ; and that 

 if the solutions were too dilute the molecular weight obtained 

 from the lowering of the freezing-point was too low, while if the 

 solutions were too concentrated, it was too high. In some cases 

 the variation of the number obtained with the concentration was 

 enormous. Prof. Carey Foster remarked that much depended 

 on the definition given of a molecule, whether it is defined as 

 that smallest quantity capable of existence per se, or as that 

 quantity which produces a given effect in depressing vapour- 

 pressure, or freezing-point, &c. Tne two magnitudes were not 

 necessarily the same. The relation observed could hardly be 

 accidental ; yet he thought that the value obtained might be a 

 quantity connected with the molecular weight but not necessarily 

 identical with it. Prof. Ramsay, in replying, said that substances 

 in dilute solutions must be regarded as in the gaseous state, their 

 molecules being so far distant from each other as not to exert 

 appreciable attraction on each other ; and as occupying but a 

 small portion of the space they inhabit. It has long been argued 

 that the molecular complexity of the gases, hydrogen, oxygen, 

 and nitrogen, must be the same, inasmuch as these elements have 

 equal coefficients of expansion within the widest limits of tem- 

 perature. A similar argument applies to substances in dilute 

 solutions ; it is much more probable that they have a simple and 

 similar m olecular structure than that the molecules, if complex, 

 dissociate to an equal extent on equal rise of temperature, or on 

 equal alteration of concentration. As regards the empirical 

 nature of Raoult's laws, it is paralleled by the empirical nature 

 of Boyle's and Gay-Lussac's laws — that is, such laws are merely 

 approximations to truth, and depend on the fact that the mole- 

 cules are sensibly beyond the sphere of each other's attraction, 

 and themselves o ccupy no appreciable space. Hence their inap- 

 jplicability athigh concentrations. — Some compounds of tribenzyl- 

 phosphine oxide, by Dr. Collie. — Contributions to our knowledge 

 of the isothiocyanates, by Dr. A. E. Dixon. — The constitution 

 ofprimuline and allied sulphur-compounds, by Mr. A. G. Green- — 

 The determination of the constitution of the heteronucleal a,3- 

 and ;33-di-derivative3 of naphthalene (second notice), by Prof. 

 H. E. Armstrong and Mr. VV. P. Wynne. Three heteronucleal 

 Kx-chloro-;3-naphthylaminesulphonic acids can be obtained when 

 I : 2-a-chloro-3-naphthylamine hydrochloride is sulphonated 

 with four times its weight of an acid containing 2 per cent, of 

 SO3. Acid No. I. is the chief product when the sulphonatioa 

 is effected at 70° during six hours ; acid No. II. is almost the 

 sole product when the sulphonation is allowed to continue for a 

 farther six hours at ioo° ; and acid No. III. is obtained, together 

 with acid No. II., when the sulphonation is effected at i6o° 

 during six hours. The determination of the constitution 

 of the heteronucleal aj3- anl y3i3-di -derivatives of naphthalene 

 is arrived at from a study of these acids in the following 

 way. Adopting the conventional symbol for naphthalene, with 

 the a- and ;3-positions indicated, and numbering the positions 

 i, 2, 3, 4, i', 2', 3', 4', as shown — 



it is obviously possible to determine the relative positions 

 ■of the three radicles, CI, NHj, and SO3H, in a hetero- 

 nucleal chloronaphthylaminesulphonic acid by determining the 

 relative positions of the three pairs of radicles, CI and NH^, 

 CI and SO3H, and NHj and SO3H, and on this result to 

 base the absolute orientation of the radicles, provided that it 

 can be shown how the radicles are situated in any one of 

 the two pairs of hetero-di-derivatives obtainable from the tri- 

 ■derivative. No absolute method, free from reproach, has yet 

 been devised for determining the constitution of any known 

 hetero di-derivative of naphthalene, but in the case of the 



hetero-oa-di-derivatives arguments may be advanced which are 

 of such weight as to leave but little room for doubt that the 

 00-di-derivatives corresponding in constitution with ^-dichloro- 

 naphthalene (m.p. = 83') have the constitution I : i' ; the 

 oa-di-derivatives corresponding in constitution with 7-dichloro- 

 naphthalene (m.p. = 107°) having then, by exclusion, the alter- 

 native formula 1:4'. The constitution of a-chloro-ySnaphthyl- 

 amine, and consequently the relative positions of the radicles CI 

 and NHg in the three sulphonic acids derived from it, was deter- 

 mined by the authors to be i : 2 (cf Nature, December 13, 1888, 

 p. 166) ; the relative positions of the radicles CI and SO3H in 

 the three acids were determined by replacing theNH2 radicle by 

 H by von Baeyer's hydrazine method, and converting the result- 

 ing heteronucleal chloronaphthalenesulphonic acids into the cor- 

 responding dichloronaphthalenes by treatment with phosphorus 

 pentachloride, and the relative positions of the radicles NHj 

 and SO3H were ascertained by replacing the CI by H by 

 reduction with sodium amalgam, and converting the resulting 

 heteronucleal naphthylaminesul phonic acids, first into the cor- 

 respondin;T chloronaphthalenesulphonic acids by Sandmeyer's 

 method, and finally into the corresponding dichloronaphthalenes. 

 Acid No. I. yields, by displacing NH.^ by H, a chloronaphtha- 

 lenesulphonic acid corresponding in constitution with 7-dichloro- 

 naphthalene, and therefore contains the radicles CI and SO3H 

 in the positions i : 4' ; it follows then, since the radicles CI and 

 NH, are in the positions i : 2, that the radicles NH2 and SO3H 

 must be in the positions 2 : 4'. The naphchylaminesulphonic 

 acid obtained from No. I. acid by reduction was found to be 

 identical with Dahl's modification of /3-naphthylamine-o-sulphonic 

 acid, and to yield a dichloronaphthalene identical with so-called 

 r;-dichloronaphthalene (m.p. = 48°), so that these and all cor- 

 responding. heteronucleal a5-di-derivatives are proved to have the 

 constitution 2 : 4'. Acid No. III., on displacing NHj by H, 

 yields a chloronaphthalenesulphonic acid corresponding in con- 

 stitution with the Badische modification of )3-naphthylamine-a- 

 sulphonicacidand with the heteronucleal a^-dichloronaphthalene 

 melting at 63°"5, and this must, by exclusion, be the i : 2'-com- 

 pound, since the only alternative formula for a heteronucleal 

 a,8-di-derivative, viz, I : 3' or 2 : 4', has been proved above to 

 belong to a;3-di-derivatives corres]3onding in constitution with 

 Tj-dichloronaphthalene (m.p. — 48°). Inasmuch, then, as acid 

 No. III. contains the radicle^ CI and SOall in the positions 

 I : 2', and the radicles CI : NH.^ in the positions i : 2, it follows 

 that the radicles NHgand SO3H must be in the positions 2 : 2'. 

 The naphthylaminesulphonic acid obtained from No. III. acid 

 by reduction was found to be identical with Bayer and Duisberg's 

 modification of /3-naphthylamine-y3-sulphonic acid, and to yield 

 a dichloronaphthalene identical with S-dichloronaphthalene 

 (m.p. = 114°), so that these and all corresponding heteronucleal 

 y83-di-derivatives are proved to have the constitution 2 : 2'. 

 Acid No. II., on displacing NH.2 by II, yields a chloro- 

 naphthalenesulphonic acid convertible into r;-dichloronaphtha- 

 lene ; it follows, then, that it contains the radicles CI and SO3H 

 in the positions i : 3', and since the radicles CI and NHg are in 

 the positions r : 2, the radicles NHj and SO3H must be in the 

 positions 2 : 3'. The authors could not, however, succeed in 

 isolating this naphthylaminesulphonic acid owing to the peculiar 

 behaviour of acid No. II. on reduction with sodium amalgam, 

 but they have been able by other methods to place it beyond 

 doubt that the acid when isolated would be found identical with 

 Brbnner's ;3-naphthylamine-/3-sulphonic acid, and would yield 

 a dichloronaphthalene identical with e-dichloronaphthalene 

 (m.p. = 135°), hence these and all corresponding heteronucleal 

 ;3,3 di-derivatives have the constitution 2:3'. These results 

 not only render it possible to determine the constitution of the 

 heteronucleal k;8- and yS^-di-derivatives of naphthalene, but also 

 afford a method of ascertaining the constitution of the two 

 hf-teronucleal sulphonic acids obtained on sulphonating i : 2 di- 

 chloronaphthalene, and of the three trichloronaphthalenes which 

 have been obtained from the three chloronaphthylamine- 

 sulphonic acids ; the method is being extended by the authors 

 to all the known chloronaphthylamines for the purpose of 

 determining the constitution of the sulphonic acids obtained 

 by them in characterizing the corresponding dichloronaph- 

 thalenes. The results, moreover, establish the correctness 

 of the opinion, long held and frequently expressed by the authors, 

 based on the higher melting-points of the "uniform" e-di- 

 derivatives {i e. di-derivatives containing two similar radicles) in 

 comparison with the isomeric 5-di-derivatives, that the e-di- 

 derivatives are symmetrically constituted ; this conclusion, in 



