266 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS pyrolysis of (+)-car-3-ene, the alcohol (formula 26) gave only meta men- thane derivatives. The structures of the menthadiene (formula 43) and its conjugated isomer (formula 44) were assigned spectroscopically and chemically. Both consumed two molar proportions of hydrogen, the former giving cis-m- menthane and the latter giving a mixture of cis- and trans-m-menthanes. It should be remembered, however, that the configurations of the menthane products of hydrogenation cannot be used as an indication of the con- figurations of the starting dienes. It has been our experience that the more stable o-, m-, or p-menthanes are obtained predominantly regardless of the configurations of the dienes. The spectra of the menthadienes (formulae 41-44 and 46) are given in Table IV and they fully substantiate the assigned structures. The con- figurations of the two alkyl groups of (--)-cis-m-mentha-4, 8-diene (formula 43) are those of (+)-cis-caran-cis-5-ol (formula 26) since the centres at C1 and C3 are not involved in the pyrolysis reaction. Although the alcohol (--)-cis-m-menth-4-en-8-ol {formula 31)may be obtained from (+)-cis-caran-cis-5-ol (formula 26) by pyrolysis it is more conveniently obtained by treating this alcohol with 0. IN hydrochloric acid at 20øC. The yield of the alcohol (formula 31) depends upon the concentra- tion of the acid. With higher concentrations of acid, dehydration and iso- merisation take place giving m-menth-3(8),4-diene (formula 44). 1('26 ) (44) (--)-cis-m-Menth-4-en-8-ol (formula 31) shows nmr resonances at ß 9.02 (d, J 5Hz, CH 3-7), 8.91 and 8.86 (singlets, CH 3- 9, 10), 8.5 (s, OH), 4.25 (s, 2H, H4, H5). There was no signal in the T 5-7.5 region, indicating the tertiary character of the alcohol. In the infra-red, peaks at 3 268 (OH), 1 642 (C----C), and 735 (cis CH-•CH) cm-1 support the structure. Con- firmation comes from its mass spectrum which shows peaks at m/e 154 (M+), 59 [(CH 3) 2COH+I, and 44 (CH •COH+). Finally dehydration of the alcohol (formula 31) with phosphoryl chloride gave a mixture of the dienes (formulae 43 and 44). (--)-trans-C aran-cis-2-ol Pyrolysis, at 410øC in the vapour phase, of (--)-trans-caran-cis-2-ol

SOME INVESTIGATIONS OF THE CHEMISTRY OF CARENE 267 (formula 45) which was obtained by the lithium aluminium hydride re- duction of (--)-trans-caran-2-one (formula 13), gave (+)-trans-p-mentha- 2,8-diene (formula 46) (85%) and (+)-p-mentha-2,4(8)-diene (formula 41, R, a- CH 3) (15%). 9 '0 ' (45) (4•) (47) OH (48) On heating (--)-trans-caran-cis-2-ol (formula 45) with a little acetic acid in a sealed tube at 210øC, (+)-trans-p-mentha-2,8-diene (formula 4t3) was again the predominant product. Its spectra (Table IV) and the fact that on hydrogenation it affords trans-p-menthane largely settle its structure. However the configuration of the 7- methyl group follows from the con- figuration of the alcohol (formula 45), since this methyl takes no part in the reaction. If it did, the pyrolysis product would be p-mentha-1,8-diene (for- mula 47). The minor product of the acid-catalysed thermal isomerisation was (+)-trans-p-menth-2-en-8-ol (formula 48) (33) an analogous alcohol to (--)-cis-m-menth-4-en-8-ol (formula 131) similarly formed from (+)-cis- caran-cis-5-ol (formula 213). The infra-red spectrum of the menthenol (for- mula 48) showed peaks at 13350 (OH), 1 1350 (C=C), and 728 (cis CH=CH) cm-1, whilst its nmr spectrum had signals at z 9.04 (d, J 13Hz, CCH3), 8.86 and 8.91 [singlets, C(CH 3) 2], 8.213 (s, OH) and 4.138 (s, H2,13). Its mass spectrum was similar to that of its isomer (formula 131) (see above). These spectra confirm its structure. Larger yields (85%) of (+)-trans-p-menth-2-en-8-ol (formula 48) were obtained when (--)-trans-caran-cis-2-ol (formula 45) was treated with 13N hydrochloric acid. a-Terpinene (formula 40) (5%) and (+)-p-menth-2,4(8)- diene (formula 41 R, a-CH 3) (15%) are other products of this reaction. CARANAMINES Only one caranamine was known, when we commenced our work (137), namely a dextrorotatory caran-4-amine (138). However during the course of the work a preparation of (+)-trans-caran-cis-2-amine, presumably Baeyer's amine (138), was described (139). We have therefore worked out (137) stereospecific routes to the following: (--)-trans-caran-cis-2-amine (for-. mula 49) (the enantiomorph of Baeyer's amine), (--)-cis-caran-trans-2- amine (formula 50), (+)-cis-caran-cis-4-amine (formula 51), (--)-cis-caran-