328 COHEN AND MARSHALL:XXXVL-The Constitution of the Arnidines. A New&kthod fay' Determining khokcukw Symmet?y.By JULIUS BEREND COHEN and JOSEPH MARSHALL.H. VON PECHNANN (Bey., 1895, 28, 869) made the observation thatthe amidine (I) prepared from benzanilide iminochloride and p-toluidinemas identical, so far as could be shown by comparison of melting pointsand other physical properties with the amidine (11) obtained frombe n zoyl-p- t oluid ide i minochloride and aniline.* *C,N,*NH-CPh :N-C,H,. C1,H,*NH*CPh:N*C7H7.The product obtained by the ethylation of this substance was shownto be a mixture of two ethyl derivatives in equal quantities, and thesewere separable by mechanical means. This fact may indicate, eitherthat the original amidine is a mixture of equal quantities of the sub-stances represented by (I) and (II), or that it has a varying structurecaused by the oscillation of the hydrogen atom marked" in theformula between the two nitrogen atoms,I n cases where the two bases differed in character from each other,von Pechmann obtained quite different results.Using ( a ) methyl-benzamide iminochloride and aniline, and ( b ) benzanilide iminochlorideand methylamine, he obtained, as before, the same product in each case,hut methylation of t h i s substance produced only one methyl derivative.This was found to be identical with the amidine obtained by combiningmethylbenzamide iminochloride with methylaniline, indicating thatthe amidine prepared either by method (a) or ( b ) had the constitutionrepresented by (111), and that a substance with the formula (IV) wasnot produced at all.CH,*N:CPh*XH*C,H,.CH,*NH*CPh:N*C,H,.(1.1 (11.1(111.) (IV-)The present investigation was undertaken with the view of deter-mining whether the introduction of optically active groups into theamidine would afford a more delicate means of distinguishing betweenthe products of the two reactions in the event of their being indistin-guishable by other means, The optically active base used in most ofthe experiments was I-menthylamine, as being easily obtainable in largequantities (Tutin and Kipping, Trans., 1904, 85, 69). The benzoylderivative of menthylamine was converted into the iminochloride andallowed t o react with the second base, and the amidine obtained inthis way was compared with that prepared from the iminochloride ofthe benzoyl derivative of the second base arid menthylamine.ThTHE CONSTITUTION OF THE AMIDINES. 329other bases used were aniline, 0-, m-, and p-toluidine, and ethylamine.It was found that in all cases the pairs of amidines had exactly thesame specific rotations, melting points, and other physical properties.I t was also shown, by ethylation of some of the amidines, that theywere single substances and not mixtures, as only one ethyl derivative wasformed in each case. These ethyl derivatives were also sgnthesised,and it was proved that the constitution of the amidine containingphenyl and menthyl groups is represented by (V), and that containingethyl and menthyl groups by (VI).C,H,*NH*CPh:N*C,,H,,.C,H,*NH* CPh : N* C,,H,, .w. 1 (VI.)These experiments confirm the observation that marked differencein character of the radicles causes the formation of one and the sameproduct in the two reactions, and there is no indication of anyequilibrium mixture, as in the case where the radicles are of the sametype. It is interesting t o note that the hydrogen atom is in all casesattached to the nitrogen atom in the less basic group.It was then thought that it would be of interest to prepare anamidine in which the two bases were enantiomorphous. I n such a case,according to von Pechmann, a mixture of two substances would beobtained of the types (VII) and (VIII).NHR*CPh:NK. NHR* CPh: NR.- + f(TII.) (VIII.)These should be produced in equal quantities, and hence any opticalactivity due to (VII) would be counterbalanced by that due t o (VIII),and the product should be inactive, This was found to be the case,the bases used being d- and I-bornylamines, which are easily obtainedby Forster’s method (Trans., 1898, 73, 390) from the correspondingcamphors.Ethylation of the amidine produced one inactive ethylderivative. It was conceivable that if the amidino were a mixtureof two substances represented by formula? (IX) and (X) :(Z)C,,H,7*NH* CPh:N*C,,Hl7(d) (I)C,,Hl7*N:CPh*N H*C,,H,,(d)(IX. ) (X. 1it might be possible by crystallising it with an optically active acid toobtain fractions which would, after removal of the acid, be active.The amidine was combined with Reychler’s d-camphorsulphonic acid,and the salt crystallised several times from hot light petroleum, Thespecific rotations of various fractions were then determined, but theywere found to be practically identical.After removal of the acid, therecovered amidine was inactive. Similar experiments were carriedout with the ethyl derivative, but no resolution of the amidine couldbe detected.It is proposed t o extend to other classes of compounds this metho330 COHEN AND MARSHALL:of examining molecular symmetry by introducing enantiomorphousradicles into different positions in the molecule. We are at presentengaged in determining the space formula of quinquevalent nitrogenby the aid of this method.EXPERIMENTAL.Phen$menthyZbenxccmidine, C,,H,,*N C( U,H,)*NH*C,H,, was pre-pared by boiling in light petroleum the iminochloride obtained from5 grams of benzanilide with 7.5 grams of Z-menthylamine.Thementhylamine hydrochloride which separated was collected, the lightpetroleum removed, and the gummy residue extracted with dilutehydrochloric acid. To the filtered acid solution, sodium hydroxidewas added, and the amicline extracted with ether. It was crystallisedfrom alcohol, and formed fine needles melting a t 110-111' :0.5072 in 20 C.C. chloroform gave, in a 2-dcm. tube, uD - 7.40";The same compound was obtained by boiling the light petroleumsolution of the iminochloride from 5 grams of benzoylmenthylamidewith 3.6 grams of aniline, The hydrochloride of the amidine whichseparated was collected, washed with a little more ether, and decom-posed with sodium hydroxide.The compound obtained after recrgs-tallisation from alcohol melted at 110-lll", and a mixture of thissubstance with that obtained in the previous experiment had the samemelting point :0.5104 in 20 C.C. chloroform gave, in a 2-dcm. tube, a,, - 7.44';0,2054 gave 15 C.C. N, a t 15' and 750 mm.The hydrochZoq*ide, prepared from both specimens, melted a t 224",and formed long needles when crystallised from dilute alcohol. It wasonly slightly soluble in cold, but fairly so in hot, water.The platinichloride was precipitated in a crystalline condition byadding platinic chloride to a dilute alcoholic solution of the hydro-chloride.whence [u]? - 146'.whence [u]F - 146'.N = 8-61.C,,H,,N, requires N = S.39 per cent.It melts at 213' :0.5504 gave 0,0998 Pt.M.W. of amidine = 333.C23H3,,N2 requires M.W. = 334.Ethyzatioiz of Pheny~menthy~benz~~a~d~~ae.2.3 Grams of the pure amidine were heated on the water-bath forten hours with an excess of ethyl iodide, the latter being then distilled.The residue, consisting of the hydriodide of the ethylated amidine,crystallised, and was decomposed with sodium hydroxide. After twTHE CONSTITUTION OF THE AMIDINES. 331crystallisations from alcohol, 1.9 grams of pure substance wereobtained in beautiful square plates, which melted at 66-67". Thesubstance was perfectly homogeneous :0.3596 in 25 C.C. chloroform gave, in a 2-dcm. tube, a, - 11-2s";0.6260 in 25 C.C.chloroform gave, in a 0.302-dcm. tube, aD - 2.95";This amidine was also produced by treating benzoylmenthylamideThe characteristic plates, melting at0.4611 in 25 C.C. chloroform gave, in a 2-dcm. tube, [a], - 14.47' ;0.2164 gave 14.55 C.C. N, at 15O and 749 mm.The hgdrochloride was not obtained in the crystalline condition, butonly as a sticky mass, soluble in water. ' The hydriodide crystallises inprisms, melting at 220'. The- platinicldoride was precipitated from anaqueous solution of the hydrochloride, and melted at 151" :whence [ u ~ ~ " -392O.whence [a]F -391".iminochloride with ethylaniline.66-67", were obtained :whence [alga - 392O.N = 7.92.C2,H,,N, requires N = 7-73 per cent.0.5756 gave 0*09SO Pt. M.W.of amidine = 368.C,,H,,N, requires M.W. = 362.Phenylmenth ylethylbenxamidine.For purposes of comparison, the amidine from benzanilide imino-chloride and ethylmenthylnmine was prepared, It crystallised fromalcohol, in which it was much less soluble than the preceding substance,in needles melting a t 1 5 7 O :0.7436 in 25 C.C. chloroform gave, in a 2-dcm. tube, a, - 3-6O;whence [a]? - 60.5".0.1645 gave 11 C.C. N, at 20" and 762 mm.The hydrocldmide was insoluble in cold water, and crystallised fromN = 7.73.C,,H,,N, requires N = 7.73 per cent.alcohol in needles melting at 28090- Z'olylmenthylbenza~aidine, C, oH,,*N: C( C,H,)*NH* C,H7.This substance, prepared from benzoyl-o-toluidide iminochloride andmenthylamine, was isolated as the hydrochloride by saturating withdry hydrogen chloride the light petroleum solution from which theprecipitated menthylamine hydrochloride formed in the reaction hadbeen separated.The precipitated gummy mass was crystallised fromalcohol and decomposed with dilute aqueous sodium hydroxide. Theamidine crystallises from alcohol in needles melting at 106-107° 332 COHEN AND MARSHALL:0.3391 in 25 C.C. chloroform gave, in a 2-dcm. tube, aD -2.85';This amidine was also prepared from benzoylmenthylamide imino-chloride and o-toluidine, the hydrochloride of the amidine separatingfrom the light petroleum almost quantitatively :0.6264 in 25 C.C. chloroform gave, in a 2-dcm. tube, aD -5.26O;0.2064 gave 13.9 C.C. N2 at 11' and 750 mm.The hydrochZoride crystallises f corn alcohol in small, rectangularThe platinichlode mas obbained as a crystalline powder :0.5654 gave 0.1005 Pt.C24H32N2 requires M.W.= 348.m-~~Zylment~~yZ6enxamidine was isolated in a similar manner to theo-tolyl derivative from benzoylmenthglamide iminochloride andm-toluidine. The base crystallises from alcohol in needles, and meltsat 89-90' :0.6268 in 25 C.C. chloroform gave, in a 2-dcm. tube, a,, - 7-13";The amidine from benzoyl-m-toluidide and menthylamine had the0.6118 in 25 C.C. chloroform gave, in a 2-dcm. tube, a, - 6.98';0.2090 gave 14.4 C.C. N, at 13.5' and 751 mm.The hydrochloride crystallises from alcohol, and is only very slightlyThe platinichlwide forms very small needles, which melt a t 217' :0.5603 gave 0.0991 Pt.C2,H,,N, requires M.W.= 348.p-Z'olyZmenthyZbenxamidine, isolated in the usual wag! from benzoyl-menthylamide iminochloride and p-toluidine, crystallised from alcoholin stout prisms, which melted at 63-68' even after severalrecrystallisations :0.6252 in 25 C.C. chloroform gave, in a 2-dcm. tube, a, - 6.55';Repeated crystallisation had no effect on this value. Whenprepared from benzoy 1-p-toluidide and menthylamine, the same ratherindefinite melting point was observed :0-6231 in 25 C.C. chloroform gave, in a 2-dcm. tube, aD - 6.54';whence [a]:' - 105O.whence [.IF - 105'.N=8*13.C,,H,,N2 requires N = 8.05 per cent.plates, melting at 218', and is scarcely soluble in water.M.W. of amidine =- 344.whence - 142'.same melting point :whence [a]F - 143'.N=8*18.C,,H,,N, requires N = 8.05 per cent.soluble in hot water. I t melts at 232'.M.W.of amidine = 346.whence [=IF - 131'.whence [a]r -131'THE CONSTITUTION OF THE AMIDIKES. 3330.2069 gave 13.9 C.C. N, a t 15" and 747 mrn.C,,H,,N, requires N = 8.05 per cent.The hydrochloride crystallised from alcohol in small needles meltingat 2 2 5 O , and the pkatinichkoride, obtained as a crystalline precipitate,melted at 208' :N = 7.94.0.2700 gave 0.0474 Pt. M.W. of amidine = 350.C2,.H,,N2 requires M.W. = 348.Menth?/ZethyZbenxamiine, C, ,H,,*N: C( C,H,)*NH4J2H5.This substance was prepared by treating tt light petroleum solutionof ethylamine (2 mols.) with the calculated quantity of benzoylmenthyl-amide iminochloride.The ethylamine hydrochloride was collected,and, after removal of the light petroleum, the gummy residue wasboiled with dilute sulphuric acid. The clear solution of the sulphatewas treated with a concentrated aqueous solution of potassium iodide,and the precipitated iodide mas separated and decomposed with sodiumhydroxide. The amidine solidified after removal of the ether used inits extraction, but all attempts at recrystallisation were unsuccessful.The substance melted at 65-67", It was purified by conversion intothe iodide, which was again decomposed. No alteration in the meltingpoint could be observed :0.6282 in 25 C.C. chloroform gave, in a 2-dcm. tube, aD - 5.63";The amidine was also prepared from ethyl benzamide irninochloride0.6240 in 25 C.C.chloroform gave, in a 2-dcm. tube, uD -5.57";0.2133 gave 16.9 C.C. N, at 13" and 750 mm.The hydrochloride was only obtained as a sticky mass, whichThe hydriodide, which was much lesswhence [u]F - 11 2'.and menthylamine, and this specimen had the same melting point :whence [.IF - 11 2O.N = 9.49.C,,H,,N, requires N = 9.80 per cent.was very soluble in water.soluble, was not obtained in a crystalline condition.M.W. of amidine = 280.The platinicliioride formed microscopic plates, melting at 218' :0.4816 gave 0.0968 Pt.C1,H,,N2 requires M.W. = 286.Four grams of this amidine were boiled for ten hours with ethyliodide, and, after distilling off the excess of alkyl iodide, the hydriodideof menth yldiethylbenxainidine which remained was recrys tallised fromwater and then converted into the amidine.On crystallisation fromalcohol, t h i s formed small, rectangular plates, melting at 31-32' :0.4756 in 25 C.C. chloroform gave, in a 2-dcm. tube, uD - 6-53';whence [u]T - 172O334 COHEN AND MARSHALL:This substance was quite homogeneous, and proved to be identicalwith the amidine prepared from benzoylmenthylamide and diethyl-amine :0.4709 in 25 C.C. chloroform gave, in a 2-dcm. tube, a,, - 6.47O ;The hydrochloride was not crystalline, but the hydriodide crystal-lised from water in fine needles, which melted at 1 5 5 O . Theplatinichloride melts a t 180'.Before the enantiomorphous bornylamines were selected for use inthe final experiments, several attempts were made to prepare otherenan tiomorphou s bases.d-Dihydrocarv ylamine was prepared From d-carvone by reductionof the oxime with sodium in alcoholic solution (Wallach, Ber., 1891,24, 3984), and the laevo-base was obtained from d-limonene by wayof the nitrosochloride and I-carvoneoxime. The benzoyl derivativeswere prepared, but it was found that during the reaction of phosphoruspentachloride with these substances, hydrochloric acid is added on tothe molecule, and hence the use of an unsaturated base was out of thequestion.As the enantiomorphous carvones are easily obtained, it wasthought that the carvomenthylamines prepared by the same methodsfrom each of these carvones would answer our purpose.Accordingly,some time was spent in attempts to obtain these bases in quantity.Baeger's method (Ber., 1893, 26, 822) was first tried, but as veryunsatisfactory yields were always obtained, the mebhod wasabandoned.Attempts were made to reduce dihydrocarvylsmine to tetrahydro-carvylamine by Sabatier and Senderens' method, but, althoughreduction took place to some extent, the reduced product was inactive.Next, 2-amino-1 -methylcycZohexane was prepared from 1-methylcyclo-hexan-2-one, obtained from o-cresol (Sabatier and Mailhe, C m p t .rend., 1905, 140, 350), and an attempt mas made to resolve thisbase by fractional crystallisation of the d-cttmphorsulphonate.Nosatisfactory results mere obtained.The active benzoylderivative melts a t 139', and the inactive substance at 144'.whence [.IF - 172'.Finally, the two bornylamines mere prepared.d- Bovny I-1-bom ylbenxnmidine.The iminochloride of benzoyl-l- bornylamide was prepared in theusual way, and treated with a light petroleum solution of d-horoyl-amine. The amidine was isolated and crystallised from alcohol, fromwhich it separated in small prisms (often in stellate clusters), meltinga t 93-94'THE CONSTITUTION OF THE AMIDINES.335The substance was optically inactive. It agreed in every respectwith the amidine from benzoyl-d-bornylamide iminochloride and2-bornylamine :0.2013 gave 12.4 C.C. N, a t 1 2 ~ 5 ~ and 731 mm.C27H4,,N2 requires N = 7.1 4 per cent.The hydrochloride crystallises from ether in long needles, whichmelt a t 297". It is very soluble in alcohol. The platinichloridemelts a t 215'.The sulphate was obtained as fine needles from dilutealcohol, and melted at 138'. The hydyiodide is insoluble in water,crystallises easily from absolute alcohol, and melts at 262".The d-camplhorsui'phonate was prepared by mixing molecularproportions of the amidine and acid dissolved in ether and alcoholrespectively. The salt was crystallised several times from lightpetroleum, and melted at 205-206" :1.0353 in 25 C.C. alcohol gave, in a 2-dcm. tube, uD + 1-34'; whenceThe salt was further recrystallised, but the melting point was notaffected, and a determination of the specific rotation gave 16.4'. Asecond crop of crystals had a slightly lower melting point, and aspecific rotation of 15.4'.A quantity of the salt was mixed with ice-water, cold ether andcold sodium hydroxide solution added, and the ethereal solution of theamidine was immediately polarimetrically examined. No rotationcould be observed,N=7.15.[a]? + 16.2'.Eth y lut i o n of d - Boi-ny l-l -born y 1 bensamidine .The amidine was heated in a sealed tube with light petroleum andexcess of ethyl iodide for eight hours in a boiling-water bath. Oncooling, the hydriodide of the ethylamidine crystallises in fine needles,melting at 257-258', and these were recrystallised from a mixtureof benzene and light petroleum and decomposed by sodium hydroxide.The amidine melted at 93-94', and a mixture of this with theoriginal amidine had a melting point of 91-93".0.1997 gave 12.4 C.C. N, at 14' and 711 mm.The hydrochloride crystallises from alcohol in needles, meltingThe d;cccmpl~olr.suZpAonnte was prepared as before and recry stallised ;0.9987 in 25 C.C. alcohol gave, in a 2-dcm. tube, uD + 1.27'; whenceThe amidine recovered from this fraction of the salt was inactivefIt was inactive :N = 6.96.C29H44N2 requires N = 6.67 per cent.at 298'it melts a t 204-206':[uyr + 16.1336 JONES : SILVER AMALGAMS.as was also that obtained from the other fractions which wereseparately examined,The authors have to thank Mr. M7. H. Perkins for help in thepreparation of the first amidine described in this paper.THE UNIVERSITY,LEEDS