PARIS E. GEORGHIOU and MARGARET A. MACDIARMID. Can. J . Chem. 58, 1759Chem. 58, (1980. In sittr preparations of 3P-tosyloxycholesta-1,5-diene and 3P-mesyloxycholesta-1.5-diene were subjected to methanolysis and hydrolysis. The products obtained were those of nucleophilic displacement with inversion at C3, a s well a s the epimers formed with retention of stereochemistry. No cyclized products could be isolated.Can. J . Chem. 58, 1759Chem. 58, (1980. Nous decrivons la preparation in situ du tosylate et du mtsylate de cholesta-1.5-diene-3P-01. L e methanolyse et I'hydrolyse en milieu acetonique aqueux deces deux composes conduit a la formation aussi bien des produits avec inversion que des produitsavec retention de configuration en position 3. On n'a pas observe experimentalement la formation d'un produit d e cyclisation.Baldwin recently proposed a set of simple empirical rules (1) to predict the relative facility of different ring closures. According to the these rules, both 3-exo-tet and 3-exo-trig ring closures leading to the formation of three-membered rings are favoured processes. The well-known "icholesteryl" solvolytic rearrangement (2) of cholesteryl tosylate 1 to form the C3-C5 bicyclo ring A steroids 2 or 3 is an example of the former type. Here, the breaking of the C3-oxygen bond of the tosylate 1 is exocyclic to the 3-membered ring formed in the reaction and the geometry of C3 is tetrahedral (see Scheme 1):There are no simple examples of the 3-exo-trig ring closure resulting in the formation of a threemembered carbon ring in steroids."'Author to whom correspondence should be addressed. 'Revision received May 12, 1980. 'Laing and Sykes (3) however, have reported that the attempted tosylation of 5a-cholest-1-en-3P-01, 4, resulted in the It appeared to us that cholesta-13-diene-3P-ol,7 could serve to ascertain whether or not a preference is exhibited for one type of three-membered ring formation over the other. In principle, the C1-C5 bicyclo ring A-type compound4 8 could be formed via a 3-exo-trig ring closure with concommitant elimination. On the other hand, the icholesteryl-type product 9 could be formed via the 3-exo-tet mode (see Scheme 2):formation of Ia,5-cyclo-5a-cholest-2-ene, 5. A mechanism was proposed which involved a 1,3-transannular hydrogen shift in the tosylate 6 which is too unstable to be isolated. The second step in the rearrangement (step ( b ) ) could be regarded a s an example of a 3-exo-trig ring closure in which C5 is an incipient carbonium ion which could be considered to be approximately trigonal:?The preparation of another Cl-C5 bicyclo ring A steroid, Ia,5-cyclo-5a-cholestan-3~-ol-6-one in which ring B is functionalized has been reported by Laing and Sykes (4). These results, however, could not be duplicated (see ref. 5).