2110 PATTEltSON AND STEVENSON : INFLUENCE O F SOLVENTSCCXXI1.-The Influence o f Solvents on the Rotation o fOptically Active Compounds. Part X VI.* TheRelationship bet ween the Chemical Constitution andthe Injuence ofa Solvent.By THOMAS STECWART PATTERSON and ELIZABETH FINDLAYSTEVENSON, M.A., B.Sc., Robert Donaldson Scholar of GlasgowUniversity.THE results which have been presented in former papers have madeit clear that the rotation of ethyl tartrate, and doubtless also ofmany other active substances, responds in a remarkable manner to* Part XV., Trans., 1909, 95, 1128ON THE ROTATION OF OPTICALLY AC'I'IVE COMPOUNDS. 2111differences in the constitution of the solvents in which the activecompound may be dissolved. Thus, in p = 5 solution in benzaldehyde,ethyl tartrate has a specific rotation of about + 4 5 O (Trans., 1909,95,3221, whilst in benzantialdoxime at about the same concentrationits specific rotation is approximately - 1 2 O , a difference of 57O.Not only can actual differences.of composition produce suchvariations in rotation, but even the subtle difference of constitutionin the syrzr and anti-oximes can readily be detected in this way,whilst other methods of investigation, absorption spectroscopy, forexample (Hartley and Dobbie, Trans., 1900, 77,509), fail to indicateany dissimilarity.It seems therefore possible that when a sufficient amount oforienting work has been done, an examination of the solvent effectof a given substance on the rotation of ethyl tartrate or others u h b l e active compound may throw valuable light on the con-stitution of the substance in question.I n our opinion the mechanism of this process is of a secondarycharacter;* that is to say, the arrangement and relationship of theatoms of the solvent molecules produce a liquid which, in the mass,has at play within it certain characteristic forces, and an activecompound dissolved in this liquid, coming under the influence ofthese forces, has its configuration and consequently its rotationaltered as the result.The change of rotation is thus only anindirect effect of the chemical constitution of the solvent. I n thefirst paper of this series the suggestion was made (Trans., 1901,79, 188) that the proximate cause of solvent influence on rotationwas that property of liquids known as internal pressure, and thatvariation of internal pressure from solvent to solvent might beindicated by variation of the solution-volume of the dissolved sub-If i t be thecase, as seems generally to be held, that absorption bands arc caused by vibrationsinside the niolecule which absorb light energy, then snch an effect wonld be of aprimary character, sincc it is directly transmitted to the medium in which it isobserved. I n regard to a iwimary effect, no two chemical compounds are likely t o beabsolutely identical, in much tlie same way that no two elements appear to 1i:ivespectral lines in commoii, but substances of analogous composition will probablyproduce closely similar effects.On thc other hand, however, two siibstances en tirelydissimilar might quite well prodnce exactly the same secondary effect, just as agiven strength of pull might be exerted either by n spring or by a system ofgyrostats, that is, in two very different ways.The boiling point of a liquidis a constitutive property of, a t least, the second ordrr, so that thc sameboiling point may be common to liquids having no chemical similarity. Themelting point of a componnd is a constitutive property of a higher order than thesecond. It is easy to see that the higher the order the more difficult it will be tointerpret the effect, and this acconnts for the striking failure to declucc anyrelationship between the simplest of physical properties, like boiling point, andespecially melting point, and chemical constitution.* As an instance of n prinia~y offect we may citc absorption spectra2112 PATTERSON AND STEVEh’SOh’ : INFLUENCE OF SOLVENTSstance. It was also suggested that, since in solutions of finiteconcentration the total volume change is doubtless shared in byboth constituents of the solution, it might be possible to correlatewith rotation values, values of molecular solution-volume calculatedfor infinite dilution, the assumption being made, as a firstapproximation, that in very dilute solution the change of volumedue to solution might be ascribed entirely to the solute.It is clearto us now, however, that this is an untenable assumption, although,of course, in some cases it may closely represent the truth, and thusaccount for certain regularities which were observed ; in general,it could only be substantiated by a strict correspondence betweenexperiment and theory, a correspondence which does not exist.Adefinite decision as to the existence of a relationship between rotationand solution-volume must be deferred until some method has beenfound of determining the true volume of the different constituentsof a solution.Some papers have recently appeared in which questions regardinginternal pressure are discussed. Winther, in a number of valuablepapers (Zeitsch. physikal. Chem., 1907, 60, 594, 651, 685), hasattempted to carry the relationship between rotation and internalpressure much further than Patterson. Dawson (this vol., p. 1041)draws the conclusion that there is no connexion between internalpressure and solution-volume, but that, nevertheless, internalpressure and rotation may be correlative.His experiments point,he considers, to the existence in solutions of compound molecules ofsolvent and solute. On the other hand, Scheuer (Zeitsch. physikal.Chem., 1910, 72, 513), setting out wit.h the view that the existenceof such compounds is highly probable, was unable to find anyevidence of their formation after a very complete and interestingexamination of the melting-point curves, the volume relationships,the viscosity, the rotation, and the dispersion of mixtures of diethyldiacetyltartrate and of menthol with various inactive solvents.Amongst other less probable hypotheses, Scheuer therefore alsosuggests, like Patterson and Winther, that rotation changes onsolution may be due to variation of internal pressure from solventto solvent.It may be added that Grossmann (Zeitsch. physikal.Chem., 1910, 73, 148), using light of various wave-lengths, hasexamined the rotation of ethyl tartrate in a number of solvents,almost all of which had already been investigated by Pattersonfor yellow light. Grossmann is of opinion that his and otherexperiments render extremely probable the existence of compoundmolecules of solute and solvent, but no effort whatever is made toshow how this conception would explain the results obtained.Other attempts to establish the existence of compound moleculeON THE ROTATION OF OP’I’ICALLY ACTIVE COMPOUNDS.2113of solvent and solute in solutions have been no more successftd.Armstrong and his students have attacked the problem in variousingenious ways (Proc. Roy. SOC., 1906, A , 78, 272; 1907, A , 79,564, etc.), as has also Philip (Trans., 1907, 91, 711), but furtherinvestigation by Usher (this vol., p. 66), and in a less degree byFindlay and Creighton (this vol., p. 536), has shown, at least, thatthe assumptions made in the particular method of attack areunwarranted.In the absence of satisfactory proof of either view, it appearsto us that the purely physical conception has the advantage ofsimplicity, and ought in the meantime to be adopted. I n seeking,therefore, to correlate the chemical constitution of a given substancewith its solvent influence, it should be carefully borne in mind thatsolvent influence being probably at least of a secondary character,any attempt to interpret it ought to be made with particular care.I n a recent paper of the present series (Trans., 1908, 03, 1838),a commencement was made in the direction indicated, when it wasshown by one of us that the solvent influence, on the rotation ofethyl tartrate, of disubstituted benzenes appeared to be governedby the general law that two substituting groups, similar or dis-similar, when in the ortho-position cause the greatest, and when inthe par&position the least, rotation in the dissolved active ester.This behaviour was observed in regard to (1) 0-, m-, and p-xylene;(2) 0-, m-, and pnitrotoluene; (3) 0- and m-dinitrobenzene; (4)a- and S-nitronaphthalene; (5) 2 : 6- and 2 : 4-dinitrotoluene.We have now examined, more or less completely, according t ocircumstances, the following solvents : phenol, anisole, phenetole,diphenyl ether ; o-nitrophenol, o-nitroanisole, o-nitrophenetole ;ncnitrophenol ; p-nitrophenol, pnitroanisole, p-nitrophenetole ;catechol, resorcinol, quinol ; pyrogallol, phloroglucinol ; a-naphthol,P-naphthol ; p-benzoquinone ; and also 0-, m-, and p-chloronitro-benzene.A considerable number of these solvents are solids, in some casesof moderately high melting point., and it was therefore occasionallyimpossible either to examine dilute * solutions at all, or to examineeven fairly concentrated solutions satisfactorily a t a low tem-perature, which renders it difficult to institute wide comparisonsbetween the data obtained for different solvents.So far as possible,however, we have extrapolated from our data to 20°, a standardtemperature adopted in previous communications, even althoughmany of the solutions could not possibly be examined in thepolarimeter at that temperature, and in spite of the fact t h a t therequisite extrapolation entails some loss of accuracy. I n other cases* With reference to ethyl tartrate2114 PATTERSON AND STEVENSON. INFLUENCE OF SOLVENTScomparison has been made at such temperatures as seemed mostconvenient.Of the solvents mentioned above, but little need be said regard-ing the three last, and we may therefore commence with them.Only one solution in each, of approximately p=25, was made upand examined in the polarimeter at several different temperatures.As in this case it is only the relative effect of the o-, m-, andp-positions that comes in question, it is unnecessary to extrapolateto 20° j numbers are given below for the interpolated temperaturegoo.Ethyl Tartrate in o-, M-, and p-Chloromitrob emzeme.Solvent. M.p. p. cc? (100 mm.). A",""."nc-Chloronitrobenzeae ..,.,..., 44'4 28-43 6'19 21 *8p-Chloronitrobenzene . . . . . , . . . 83 '0 24'26 3 '8 15'7* R,= 100 aD/p ; it does not differ much from specific rotation.It is clear that in this set of compounds, also, a,s in the othersalready mentioned, the presence of the two substituents in theo-position brings about a much more powerful solvent influencethan in either of the other two positions.A general idea of the behaviour of most of the remaining solventsexamined will be obtained from the diagram, which represents theeffect of diluting ethyl tartrate with the various inactive compoundsmentioned at a temperature of 20°, subject to the limitationsreferred to above.Phenol (m.p. 42*5°).-Solution in phenol very greatly modifiesthe specific rotation of ethyl tartrate. The value rises from +7*S0in the pure ester to +48.5O at infinite dilution in phenol, anincrease of 40.8O. Since benzene itself has but little effect (Trans.,1902, 81, 1098), it is clear that the introduction of the hydroxylgroup must profoundly modify the interatomic forces of themolecule.Anisole.-The replacement of the hydroxylic hydrogen by amethyl group brings with it a very great change in solvent influence.The specific rotation of the ethyl tartrate drops to +6*8O in ap=25 solut'ion, that is, to a somewhat lower value than in thehomogeneous ester.It will be noticed, too, that the shape of theconcentration-rotation curve is of the opposite type to that of phenol,being concave to the point of origin of the diagram, in such a waythat, starting from the value for the pure ester, the rotationincreases slight.$ on dilution with anisole to reach a maximumvalue of about + 8*2O between p=50 and 60, and then diminishesagain fairly rapidly to about +2*5O at infinite dilution.€'henetole.-In comparison with the difference between phenoland anisole, that between anisole and phenetole is slight.Foro-Chloronitrobenzene .. . . . . ... 32'5" 25.733 + 7-26 + 28 '2ON THE ROTATION OF OPTICALLY ACTIVE COMPOUNDS. 2115p = 25 solutions the rotation values are much alike (anisole, + 6.8O ;phenetole, + 7 - 3 2 O ) , but it will be noticed that in regard t o the formFIG. 1.Coilcentration-rotat~on CUTVCS for ethyl tartrate i n various solvents.+ 50"45403530 s ' 25*@ ulaG .@&2015105I ---+--- 3---20 40 60 80 100Concentration (p).of the concentration-rotation curve these two solvents differ, andat infinite dilution the values for the rotation of the dissolvedester are much more divergent (+ 2 ' 5 O and + 7-3O respectively)2116 PATTERSON AND STEVENSON : INFLUENCE OF SOLVENTSThe curve for phenetole is of the same type as that for phenol;there is a minimum rotation of + 7O at p=50, after which therotation rises to reach the value +8O at infinite dilution.Biphenyl Ether (m.p. 28O).-Only one solution was examined indiphenyl ether, of p=24*58. Its observed rotation at 20° was1-95', which, assuming a density of unity, gives a specific rotationof + 7 * 9 3 O , about one degree higher than that in anisole andphenetole at the same temperature and concentration.These results seem to make clear the fact that the powerfulinfluence of phenol in this direction is t o be attributed neither tothe oxygen atom of the molecule nor to the phenyl group or-inthe case of diphenyl ether-groups, but rather to the presence ofthe hydrogen atom of the hydroxyl group.o-Nitrophenol (m.p. 45O).--Since an o-nitro-group in toluene andother monosubstituted benzenes produces a much greater increaseof solvent influence than a m- or a pnitro-group, we had expectedthat o-nitrophenol would have a greater influence than phenol. Onthe contrary, however, the rotation at infinite dilution (+ 1 7 O ) ,although considerably higher than that of homogeneous ethyltartrate, is much below the rotation in phenol. The effect ofo-nitrophenol in this respect is in no way the mean of the effectsof nitrobenzene and phenol, which are both above +40°. Theinfluence of the substituents is thus certainly not sdditive.The concentration-rotation curve for o-nitrophenol is a straightline or nearly so.o-Nitroanisole.-The exchange of the hydroxylic hydrogen atomof o-nitrophenol for a methyl group brings about, a t infinite dilution,a considerable in~rease-l9*5~-in the rotation of the dissolvedester, namely, from + 1 F to + 3 6 * 5 O .o-flitrophene t ote .-In o-nitrop henetole the concentration-rotationcurve is very similar to that for the corresponding methyl ether,but lies wholly below it.A t infinite dilution the rotation wouldbe +27O.There is thus between phenol and its ethers, on the one hand,and o-nitrophenol and its ethers on the other, a relationship of aninverse character when the solvent effects at infinite dilution areconsidered. I n the former, the high rotation brought about byphenol gives place to a low rotation in anisole, and rises againsomewhat in phenetole, whilst in the latter compounds the com-paratively low value in o-nitrophenol rises to a fairly high value ino-nitroanisole, t o fall again considerably in o-nitrophenetole.p-Nitrophenol (m.p. 114°).-Owing to the higher melting pointsof the paraderivatives of phenol, observations could not be made,in this series, in so complete a fashion as in others, but sufficienON THE ROThTlON OF OPTICALLY ACTIVE COMPOUNDS. 2117data have been obtained to elucidate the general behaviour of thecompounds concerned. In pnitrophenol, two solutions wereexamined in regard to rotation, density determinations, however,being omitted. In a solution of p=47*86 by interpolation from theobserved data, IZl,”” = 34*6O,* whilst for p = 24.98, R1,G’’ = 46O, whenceat infinite dilution, also f o r looo, the value of R , would beapproximately = 60°.These are very high values, and they increaseas the temperature falls, so that a t ZOO the specific rotation, whichin this case would have a slightly lower value than R,, of aninfinitely dilute solution of ethyl tartrate in p-nitrophenol wouldprobably lie a t about + 75O. p-Nitrophenol is thus considerablymore powerful even than a-nitronaphthalene or o-dinitrobenzene,the most powerful of the solvents hitherto exa.mined. But owingto the extensive extrapolation requisite in this case, the concentration-rotation curve for pnitrophenol can only be regarded as a somewhatrough approximation, for which reason it is shown as a broken linein the diagram.p-Nitroanisole (m.p. 54O).-This solvent on mixture with ethyltartrate causes a, gradual and comparatively slight increase ofspecific rotation. For a solution of p = 53.84, CUE = + 12.5O. Byextrapolation of the observed readings for a, p=26*08 solution, thevalue a2,o (100 mm.)= +5-lo is fonnd. Assuming a density of1.2, which would certainly not be far from the truth, the specificrotation at infinite dilution would have it value of, approximately,+ 2 1 O .p-Nitrophenetole (m. p. 60°).-Solutions of p = 49.66 andp=24*37 were examined in this solvent. The specific rotations ofthese at 20°, assuming a density of 1.2 in each case, would benearly + So and + loo. With increasing dilution, therefore, therotation of the dissolved ethyl tartrate increases, although onlyslowly, to reach a value of about + 1 3 O at infinite dilution.m-Nitrophenol (m.p. 96O).-Only one solution-p = 49.76-wasexamined. Its behaviour is referred to below.2 : 4-Dinitrophenol (m. p. 114*5O).-We also examined one solu-tion (p= 75.15) in this substance. Its rotation was somewhat lowerthan that of an equally concentrated solution in o-nitrophenol.General Discussion of the Foregoing Results.1. We may examine first the solvent influence of phenol and itsthree nitrederivatives, which will be best accomplished by con-sidering the observed rotations for p=50 solutions at a temperatureof 70°:* See note on p. 21132118 PATTERSON AND STEVENSON : 1NFLUENCE OF SOLVENTSa j+; B d BuzPhenol + 45.5"Anisole + 2.5Solvent.P. a'," (100 mm. ). Difference.......... 2 *5"3 52 -4o-Nitrophenol 51 '13 + 9.3" .................. Phenol 48 '2 11 *8nt-Nitrophenol 49'76 15.3y-Nitrophenol ......... 47-86 17.7.........The influence of phenol is diminished 2 ' 6 O by the introductionof a nitro-group in the o-position; it is raised 3 . 5 O by a nitro-groupin the m-position and by a further 2*4* if the nitro-group be in thep-position, a behaviour which is thus just the opposite of thatobserved in other cases of ortho-, meta-, and para-isomerism.2. Comparing the behaviour of phenol, o-nitrophenol, and p-nitro-phenol with that of their respective ethers, it is to be noticed thatthe very great solvent influence of phenol disappears almost entirelyin anisole, phenetole, and diphenyl ether, and that in a similarmanner the very powerful effect of p-nitrophenol is greatlydiminished in its methyl and ethyl ethers, and by an approximatelyequal amount in the two cases, some 50°.On the contrary, how-ever, the comparatively feeble effect of o-nitrophenol is quite con-siderably raised in its ethers. Thus phenol and p-nitrophenolappear t o behave in an analogous manner, and to differ fromo-nitrophenol, as is shown in the table below.4i dG $4 2 *- Z d B aa*a.2 : 9 Ba'Z .P 2PH P T 4 2 0u1 u1 HG 6 Ap-Nitropherol + 75"p-Nitroanisole +21 +54" 1 o-Nitroanisde +36'5 - '"'" o-Nitrophenol + 17"3. As shown in the following table, o-nitroanisole has a greatersolvent influence than pnitroanisole, and o-nitrophenetole a greatereffect than p-nitrophenetole, and by almost the same amount.Rotation of Ethyl Tartrate.[aIF [slyInfinite I n fi 11 it eSolvent.dilution. Difference. Solvent. Dilution. Difference.o-Nitroanisole ,., 1-36-8" 15.80 I o-Nitrophenetole ... + 27" 140p-Nitroanisole ... 21 *O p-Nitrophenetole ... 13Although, theref ore, the behaviour of the nitrophenols is unusual,inasmuch it9 the two substituting groups produce the greatestsolvent effect when in the pposition, and least when in the o-position,this exceptional behaviour does' not extend to their ethers, whichexhibit the regularity previously described for the chloronitro-benzenes and a number of other substancesON THE ROTATION OF OPTICALLY ACTIVE COMPOUNDS.21194. It is of interest to compare, so far as is possible, our presentresults with some obtained by other methods having a similar aim.E. C. C. Baly and Miss Ewbank (Trans., 1905, 87, 1315) havedescribed the absorption curves for phenol, anisole, and phenetole.They point out that the curves for the ethers are identical, andthat they differ in one particular * from that of phenol itself.This agrees roughly with what we have found, for, although theconcentration-rotation curves for anisole, phenetole, and diphenylether are certainly not identical, still, in respect to the magnitudeof their effect when compared with phenol, they are much alike.That for phenol differs very markedly.We had also, during the course of this research, compared ourresults with the absorption curves given by Baly, Edwards, andStewart (Trans., 1906, 89, 512) for the nitrophenols and theirethers, but in a more recent paper Baly, Tuck and Marsden (thisvol., p.571) have explained the anomalous character of one earlierexperimental result -f as being due to solvent influence, and haveconsiderably altered the theoretical views formerly held. They haverejected the idea of the existence of a quinonoid structure, not onlyin the free nitrophenols, but also in their alkali salts, and theydraw the conclusion that o-nitrophenol and o-nitroanisole aresimilarly constituted, the same holding for the met& and para-compounds.It may be pointed out that our results, if it be legitimate tocompare them with those of Baly, are not in agreement wit,h thisconclusion, since, as stated above, phenol and p-nitrophenol arerelated to their ethers in a similar manner, the solvent influence oftho phenols being much greater than that of the ethers, whereasthe opposite is the case for o-nitrophenol and its ethers.Theethers of phenol, o-nitrophenol, and p-nitrophenol, however, behavenormally amongst themselves. It might therefore be concludedthat phenol and pnitrophenol are similar in structure and differentfrom o-nitrophenol, but that the ethers are all of analogousconstitution.The data which we have obtained in the examination of a numberof polyhydroxy-benzenes and of p-benzoquinone may be dealt withvery briefly. I n the table below, there are quoted interpolatedvalues for the observed rotation at looo of solutions all of aboutthe same concentration.this difference.* We are not aware what degree of importance Professor Baly would attribute tof That the spectrum of o-nitroanisole differed from that of o-nitrophenol2 120 PATTERSON AND STEVENSON : INFLUENCE OF SOLVENTSRotation of Ethyl Tartrate.Benzene ..................75.199 + 10*9=*Phenol ..................... 74-39 13.82Catechol .................. 74'81 13.45Resorcinol., ................ 74 -10 16 *55Quinol ..................... 74.47 17.60Pyrogallol.. ................ 74 '93 13.25Phloroglucinol ............ 74 *33 17 '20p-Benzoquinone ......... 74 -98 10.90Solvent. P. a;''''' (1 00 mm.).* By extrapolation from figures given in Trans., 1902, 81, 116.On account of the rather high melting points of these substances,only one concentrated solution in each was examined.Cat ec hol, Resorcinol, Quino1.-The first hydroxyl group intro-duced into the benzene ring produces a considerable increase ofsolvent influence.A second, however, in the o-position t o the first,instead of causing a further increase, brings about a slightdiminution of less than half a degree. I n the m-position, on theother hand, an increase of nearly three degrees results, whilst inthe p-position the increase is almost four degrees.Pyrogallol, Phloroglucino1.-The behaviour of these two solventsis in remarkably close accordance with what might now beexpected. The three hydroxyl groups in the vicinal position inpyrogallol bring about a small diminution its compared with catechol,whilst in the mposition in phloroglucinol there is a considerableincrease as compared with resorcinol. It is thus quite clear thattwo hydroxyl groups have least effect in the o-position and mostin the p-position, a behaviour which is thus similar to that shownby a nitregroup and a hydroxyl group when present together inthe benzene ring, but opposite to that which is characteristic of twonitro-groups, two methyl groups, or a methyl group and a nitro-group.p-Benzopuinone (m.p. 166°).-p-Benzoquinone, in a solution ofapproximately the same strength as for the polyhydroxy-benzenes,caused for u1,Ooc the value + 1 0 * 9 O , a considerably lower rotationthan is produced by phenol or catechol.It might therefore beargued that a quinonoid structure of the solvent causes, in ethyltartrate, it lower rotation than a simple phenolic constitution, andthat therefore the low rotation produced by o-nitrophenol as com-pared with phenol and p-nitrophenol may be due to it quinonoidstructure for the o-nitrophenol. We would not venture, of course,on this slight evidence, to draw any definite conclusion, but thefact is worthy of consideration along with others bearing on thequestion.a- and &Naphthol (m. p. 94O and 122O).-We also examined onON THE ROTATION OF OPTICALLY ACTIVE COMPOUNDS. 2121solution each in a- and &naphthol. In the former, REoo (p=25*24) =+ 2 9 . 7 O , and in the latter, &Eoo ( p = 29-01> = + 43'8O.The differenceis thus very considerable, and since the a-compound may be regardedas an o-hydroxy- and the &compound as a m-hydroxy-derivativeof benzene, the behaviour of these two solvents is strictly inaccordance with that of the other phenols dealt with above.Influence of Change of Temperature.With regard to t,he influence of change of temperature on therotation of these solutions, but little need be said, since the behaviourFIG. 2.Ternperchre-rotation cwrz'es for ethyl tartmle in various solvents( p = approximately 75 in each case).II I20" 40" 60" 80" 100" 120" 140"Ten iperature.observed has been entirely in agreement with what has been dis-covered in other cases. The effect of rise or fall of temperature isa function of the value of the rotation at the temperature chosen.I f the rotation at To be above a certain critical value for thattemperature, a, value which can be fairly definitely stated,* thenheating above TO will cause fall of rotation, whereas if the rotationbe below the critical value, further heating will bring about increaseof rotation. The temperature-rotation curves in Fig.2 illustrate* For 20°, for example, this critical value for specific rotation lies about3.18" to +20°.VOL. XCVII. 6 2122 PATTERSON AND STEVENSON : INFLUENCE OF SOLVENT3this point. The rate at which the rotation increases with rise oftemperature in pyrogallol (j = 74.93) is greater than that in catecholsolution of about the same concentration, in agreemeht with thefact that the rotation in the former solution is less than in thelatter.In phenol, with its greater solvent influence, the rate ofincrease of rotation is less than in catechol. I n resorcinol there isat first an increase of rotation and then a diminution, a maximumrotation occurring at a temperature of about 60°. This maximumrotation is therefore the critical value at 60°. Our data forphloroglucinol are rather scanty, but it is practically certain thatFIU. 3.Temperature-rotation cwves for ethyl tartrate in various solvents.-1--3-20" 40" 60" 80" 100" 120" 140"Temperalure.this solvent would show a similar behaviour, and therefore that amaximum rotation probably occurs at a lower temperature than60°.I n quinol only a fall of rotation was observed, but there isdoubtless a, maximum rotation somewhere about the temperature40°, which would have a greater value than the maximum inphloroglucinol or in resorcinol. The critical value of the rotationis higher the lower the temperature at which it occurs.The above remarks apply equally to the behaviour of a- and&naphthol and the nitrophenols. In phenol itself it may be notedthat whilst the temperature-rotation curves for solutions of highconcentration are concave to the point of origin of the diagramON THE ROTATION OF OPTICALLY ACTIVE COMPOUNDS. 2123that for a p=50 solution is almost a straight line but with a slightconvexity, which, as the solutions are diluted, becomes graduallymore pronounced.This resembles the behaviour of quinoline(Trans., 1909, 95, 323), and it is, of course, possible that others ofthe solvents we have dealt with would show the same behaviour insolutions more dilute than those we have investigated.The experiments recorded in this paper seem, we think, to justifythe conclusion that the chemical constitution of a solvent is notmerely reflected-which, of course, it must be-but is reflected in acomparatively simple manner, in the influence of that solvent onthe rotation of dissolved ethyl tartrate, and that, conversely, thismethod might be used for the investigation of chemical constitution.The method differs from that of refractive index and of magneticrotation, inasmuch as it does not seem possible to calculate anyconstant increment or decrement for a given group of atoms, for asingle atom or for a difference in the linking of an atom in themolecule, but these are certainly indicated qualitatively.There isperhaps a closer connexion between the phenomena we havedescribed and those of absorption spectroscopy. So far as we areable to judge, knowing only one of these methods intimately, thesolvent-influence of a compound on the rotation of ethyl tartrateaffords a more delicate criterion of chemical constitution thaneither of the others, as witness, for example, the effect of orthe,meta-, and para-isomerism, the difference between phenol andanisole, and especially the difference between the syn- and anti-oximes. We hope to describe the further study of the subject infuture papers.EXPERIMENTAL.Ethyl Tartrate in Various Solvents.dhloronitrob enzene.p=25'73:t ..................24.7" 30.55" 40.55" 49.2" 58" 82.9Oa: (100 mm.) ... 7.7 7'654 7'638 7.61 7.5 7'33m-Ciiloronitro b enzene.t .................. 36.9" 62'2" 77.4" 89.7" 1062"aD (100 mm.) ... 5.754 5.9 6.125 6'046 6,236p = 28.43 :p-Chloronitrobcnzene.t .................. 82.1" 105-2" 117" 128.2" 147'1"qt (100 mm.) ... 3.i35 3.885 4.098 4.106 4.12p = 24-26 :6 2 2124 PATTERSON AND STEVENSON : INFLUENCE OF SOLVENTEthyl Tartrate in Various Solvents (continued).Ph enot,t 20" ao.75" 62-30 69'5" 92.4"a: (100 mm.) ... - +5%95 +4.885 +4*525 4 *085[a]: ............... +385 +35*47 +80'96 -:-29'10 26-86I. p=14*79:..................11.p=23*85 :t .................. 18.8" 20" 32.4" 41.9" 53.5" 65.5 96.9" 130.4'a2(100 mm.) ... +9-039 - 8.445 8.031 7.586 7'23 6.345 5.666[a]: ............... 34-25 $34.2 32.32 30.98 29-55 28'42 25.7 23-65111. ~'48.2:t .................. 19" 20" 39" 65-2" 92.3" 140.6"a: (100 mm.) ... +13*124 - 12.64 11-95 11.27 10.15[a]: ............... 23.99 23.94 23'44 22.62 21.83 20.54IV. p=64.78:t .....................a: (100 mm.) ......[UJE, ..................v. p = 74-39 :E .....................a: (100 mm.) ......,a]; ..................VI. p = 79.24 :t .....................u: (100 mm.) ......[a]; ..................I. +-t. d.45'4 1.072566.4 1 -053978.8 1.043138.5" 1.0787I v.&t. d.14'6" 1-166535'5 1.145165 8 1'126183'4 1,099720" 20.9" 64.9" 100*3" 126"- 12,853 13.309 13.058 12.717.05 17'10 18.38 18'78 18'5218-75" 20" 53.5" 76.5" 89.7"11.99 - 13.225 13.62 135'7513'78 13.82 15,60 16.38 16.7619.95" 20" 55.5" 87'1" 117'2"11.485 11.534 13'248 14'018 14.20912.30 12.37 14'62 15-90 16.62Densities Determined.IJ.- t. d.21.4" 1.105037-9 1.090368.2 1.072178.1 1.053899.8 1.0332v.&t. d.16.1" 1-1 75528.7 1 -163542.0 1-149964.7 1.1284111. - t. d.11 '9" 1.14041'1252 33.546.5 1.113374.6 1.087899.2 1.0640v I.w.t. d.13.5" 1.183738 '4 1-160658.4 1.141282 1 1.117ON THE ROTATION OF OPTICALLY ACTIVE COMPOUNDS. 2125Ethyl Tartrate in Various Solvents (continued).A niso l e.I. p=9*99:t ........................18.2" 20" 27 *3" 35.2"at (100 m.) ...... + 0'432 - + 0.536 + 0.622[a]; .................. + 4'27 +4'5 + 5.34 + 6-2511. p=24-62 :t ................... 14.6" 20" 33.2" 46.8" 52-9" 68.9" 72-8"............... + 6'24 6.8 8'20 9.75 10.15 12.11 12.41aE, (100mm.) ... + 1'602 -- 2.068 2.428 2.512 2.952 3.012111. p = 49.85 :t ....................... 19" 20"- a; (100 mm.) ...... 4'324[a]:. .................... 7 '96 8-1I.-7 t. d.17.5" 1.014225.4 1.0064I. p=9.99:t... ...................a: (100 mm.) ......[a];. .................11. p=24'96:t .....................[a]:. ................. a: (100 mm.) ...111. p = 51-73 :Densities Determined.11. * t. d.14.9" 1.042432.5 1.026145.5 1'013055.5 1.0033Phenetole.195" 20": 22-1" + 0.76 - 0.786 + 7.72 7.75 8.0019.2" 20" 31.6 + 1'832 - 2.202 + 7-23 7'32 a*f9t........................ 1 a -20 20"[a]: .................. + 6-72 + 6'86- a; (LOO mm.) ...... + 3.742Densities Determined.11. - t. d.17.9" 1.016128'1 1.007138.3 0.996948.4 o m 6 930.6"4.9049 *12111. - t. d.15.8" 1'093325'6 1.0835- -43.5 50'510.14 10'922-51 2.68631-0"4.4328.02126 PATTERSON AND STEVENSON : INFLUENCE OF SOLVENTSEthyl Tartrate dn Various Solvents (continued).Diphenyl Ether.t. ..................... 18.7" 20" 51.8" 66.0" 76.3" ...... 3 2 8 8 a4, (100 mm.) + 1.824 1.95 2-68 2-92p = 24-58 :*Nitro pheno 1.t . ........................ 60.1" 81.4" 112.2"I. p=23.89 :......... a: (100 mm.) + 4.82 4-972 5-007[a]: +16*20 16.98 17.66 .....................11. p=51*12 :t ......................... 56.8" 78'2" 97.3"a: (100 mm.) ......... + 9-015 9.486 9.766[a]; ..................... -t14*31 15'35 16.08111. p = 74.95 :t ........................ 16" 43.4" 87.3' 115.1"[a]: + 9.74 12.14 14.38 14.21u4, (100 mm.) ...... f9'142 11.119 12-66 13.056.....................Densities Determined.I. IJ.* 111.+7 7-y - t. d. t. d. t. d.47.7" 1.2603 56.8" 1.2323 12.5" 1.255865.9 1.2416 78.2 1.2086 27.1 1.238578.1 1.2295 97.3 i m 7 8 49'4 1-2164 - - 72.5 1.1910* By interpolation from the determinations for the other two solutions.-o-Nitroanisole.t ................... 17.9" 20" 345" 43.3" 53.2" 62.8"I. p=9*79 :aE, (100mm.) ... + 3.748 - 3.548 3-490 3.354 3'26[a]; ...............+ 3 0 9 30.6 29'5 29.2 28 3 27.711. p=21*17:t. .........."...... 17.4" 20" 29.1" 39'3" 50" 55'8"at (100 nim.) ... 6.666 - 6.604 6-472 5.442 '324[a]: ............... 25.35 25'3 25'4 25.1 25'2 24.8Densities Determhed.I. 11.& &t. d. t. d.17.7" 1.2471 14.9" 1'244233 1 -231 9 35 1 *223742'9 1'2219 53'2 1.205359.7 1 '2050 66-4 1'191ON THE ROTATION OF OPTICALLY ACTIVE COMPOUNDS. 2127a; (100 mm.) + 11262 10.646Ethyl Tartrate in Various Solvents (continued),o-Nitro phene t ole.t. ................. 18-4" 20" 30*6" 53.3" 65.5"uk (100 mm.) ... + 5-698 - 5.812 5'906 5'896[a]; ............... +19*11 19.2 19.7 20.41 20'6I. p=25'09 :11. p=39-43:t ................... 15.9" 20" 25.1" 29.1" 42"aI, (100 mm.) ... + 7'252 - 7'544 7'57 7'848[a]: ............. + 15 '41 15 '75 26-16 16.27 17.06+ 17'624 16.878 16'274Densities Determined.I. 11. - - t. d. t. d.17.1" 1.1896 14'6" 1.194833'4 1,1734 25'9 1'183340'4 1.1659 33.3 1.175754'3 1.1524 41 1'1680p-Nitroanisote.aE, (100 mm.) ...... + 6,082 6.25 6.515 6,075 6 078p = 26.08 :t. ..................... 52.0" 55.8" 64.8" 73.6" 77"p = 53-84 :t :....... .......... 20" 35'1" 44'4" 51.3" 57'6"[a]: ............... +12'5 +13'63 14.23 14.58 14.91a; (100 mm.) ... - +8.876 9'198 9'376 9.54Densities Determined.t 35'1" 4 1 -4" 48" 57.2"d 1'2101 1.2038 1'1974 1.1883p-Nitrophenetole.t 50.1" 57'3" 67.3" 82'5" 49'4" 54-4" 66'6" 69.7"a; (100 mm.) ... +3'88 3'995 4-21 4'345 1-6.475 6.73 7-175 7.28p =49*66 : I ..................p = 2 4 * 3 7 :m-ATitropherroZ.t .................... 35.4" 41'4" 50.3" 57.7"a; (100 mm.) ...... +I53 15.68 15.665 15 55p =49*76 21 28 ROTATION OF OPTICALLY ACTIVE COMPOUNDS.Ethyl Tartrate in Various Solvents (continued),2 : 4-Dinitroyhenol (m. p. 114*5O).t ........................ 84.5" 107" 121.5"at' (100 mm.) ...... + 10'696 11'2i2 11.564p = 75.15 :Cat echol.p = 74.81 :t ..................... 41.4" 68.9"aE, (100 mm.) ..... +12.12 13.04PE eso r cino 1.t ..................... 15.9" 479"a; (100 mm.) ...... +16'25 16-86p = 74.12 :Quin ot.p = 74.47 :t ........................ 70.3"a; (100 mm.) ......... -t.18'1Pyrogallot.p = 74.93 :t ........................ 49.5"a: (100 mm.) ......... 11.88494.3" 103.7"13'40 13.5374.3"16-89105"17.578 '9"12.868Phloroglucinot.p = 74-33 :t ........................ 69.5" 93.6"(100 mm.) ......... +17.48 17.43a-iVap7hthol.p = 25.24 :t ........................ 105%"aI, (100 mm.) ......... +7*276P-ATaphthol.p = 29.01 :t ........................ 132.7"a: (100 mm.) ......... + 10'92p-Benzo puinone.p=74*98:t ........................ 87-5" 115"a', (100 mm.) ......... + 10-804 10'968TEE UNIVEILSITY,-~ GLASGOW.104.2"16'644136.5"16-53108.6"13.344104*1"17-10121 -5"6.732150" + 9 988122'9"13'12