SOME INVESTIGATIONS OF THE CHEMISTRY OF CARENE 273 affected by base, thus confirming the trans- arrangement of groups, but it was reconverted to the racemic form of the alcohol (formula t31) with lithium aluminium hydride. A racemic form of the alcohol (formula 131) having an unspecified configuration at C3 has been obtained (50) by treatment of car-3-ene dibromide with silver oxide, giving the ketone (formula t32 configuration at C3 unspecified) followed by lithium aluminium hydride reduction of the latter. The configuration at the carbinol centre of our alcohol (formula t31) was derived by the use of Horeau's reaction (20). The other ring rearranged deamination product was the very volatile hydrocarbon (formula 133), namely (+)-3•,7,7-trimethyl-anti-cis-anti- tricycloI4,1,0,02,4]heptane. In its nmr spectrum it showed no olefinic hydrogen or vinylic methyl group resonances, but it displayed a methyl doublet at z 0.05, two methyl singlets at 8.07 and 0.11, a saturated methylene signal at 8.21 (2H) and cyclopropyl protons (SH) at 8.01-0.135. Its mass spectrum displayed ions at m/e 13t3 (M+), 107 (M +-- CH3CH2), and 03[M +--- (CH 3) 2CH] (base peak). The loss of CH 3CH2 and (CH 3) 2 CH from the substituted cyclopropane rings is characteristic of these groups and, together with the nmr spectrum, these features establish the struc- ture of the tricyclic hydrocarbon (formula 133). hN02• NH2 (•2a) H''O•H HO H CH• CH 3 The formation of these rearrangement products may be portrayed as above, using the diazonium ion from (--)-cis-caran-trans-4-amine (formula 52a) as example. In the example quoted, the reaction may be synchronous but in the case of the epimeric 4-amine, the geometry does not pertain. There is no

274 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS other evidence to support a concerted reaction in the initial stages. However in the formation of the alcohol (formula 61), rearward attack of water on the carbonium ion at C3 must be synchronous xvith the breaking of the C2-C3 bond so as to afford the R- configuration at the chiral centre. OXIDATION The oxidation of (+)-car-3-ene (2) at its allylic centres with gaseous oxygen has been the subject of recent investigation (18). The products were (+)-car-2-en-trans-4-ol (formula 64), (--)-car-3(lO)-en-trans-4-ol (formula 65), and (--)-cis-car-4-en-trans-3-ol (formula 66). Earlier Semmler and Schiller (51) isolated what was believed to be car-3-en-5-one (formula 67) from the oxidation products of (+)-car-3-ene with permanganate. This unsaturated ketone {formula 67) was regarded (52) as a very unstable entity and its isomer car-3-en-2-one (formula 68) as being incapable of separate existence because of its ready isomerisation into eucarvone (formula 69). Initially we decided to reinvestigate the oxidation of (+)-car-3-ene (for- mula 2)using (a) potassium permanganate (53) and, (b) chromium trioxide (54). When (+)-car-3-ene was stirred in acetone at room temper- ature with potassium permanganate, a complex mixture was obtained, 10% of which consisted of neutral volatile oxidation products, in addition to recovered (+)-car-3-ene. After distillation, the volatiles were separated by a combination of column chromatography on silica-gel and preparative glc on Carbowax columns. The principal oxidation product was (--)-car- 3-en-5-one, having the absolute configuration shown in (formula 67) and this represented 75% of the total neutral volatile oxidation products*. Its *The yield of each oxidation product is expressed as a percentage of the total volatile oxida- t ion products.