Acetolysis of the toluene-p-sulphonate ester of 2-(2-hydroxyethyl) -6,6-dimethylbicyclo[3.1.13 hept-2-ene (nopol) (1; R = OH) gave a good yield of the acetate of the previously unknown tricyclic pinane derivative 8,8-dimethyltricyclo[5.1.1 .02*5] nonan-2P-01 (4a). This molecule has a bridgehead hydroxy group, and so is remarkably stable to acids, despite being highly strained from having two cyclobutane rings. However, the hydroxy group is adjacent to the new cyclobutane ring, so that (4a) is readily oxidised to the gcetate of the hydroxytetrahydrofuran, 2phydroxy-8,8-dimethyl-l O-oxatricyclo-15.1.1 -1 29s]decane (5; R = OAc). Hydrolysis of the acetate group, followed by oxidation of the alcohol (5; R = OH), yields 8,8-dimethylbicyclo[5.1.1] nonane-2,5-dione (6). The base-catalysed dehydration of 2-(2-hydroxyethyl)-6,6dimethylbicyclo[3.1.l]hept-2-ene (nopol) (1; R = OH) yields a mixture of the dienes (2) and (3), but the reaction does not yield any tricyclic pinane derivatives., The Clarke-Eschweiler cyclisation has been demonstrated3 to take place with 2-(2-aminoethyl)-6,6-dimethylbicyclo[3.l.l]hept-2-ene (nopyl- amine) (1; R = NH,), but does not yield a system with a pinane skeleton; rearrangement gives products based on the fenchyl and terpinyl systems. The only reported tricyclic pinane derivative is 1 O,lO-dimethyltricyclo~7.1. 1.02*7]undec-2( 7)-en-6-one, prepared from P-pinene during the synthesis of P-~elinene.~ In this case, a second six-membered ring was added to the pinane skeleton. This suggested that a pinane with two cyclobutane rings may well exist, but could be highly labile, and thus isolable only under mild conditions. We attempted to prepare it by the solvolysis of the toluene-p-sulphonate of nopol. Results and DiscussionThe toluene-p-sulphonate of nopol (1; R = OH) was prepared by conventional methods. It readily underwent solvolysis in