Ultraviolet irradiation of 3P-nitrito-5a-lanostane was found to follow an unusual reaction pathway and yield 2-oximo-3-oxo-3,4-seco-5a-lanostane (IV). The new aldehyde was easily converted to oxime and phenylhydrazone derivatives. Acetylating the photolysis product led to an oxime acetate (IX) which readily rearranged to 1-cyano-3,4seco-5a-lanostane (111). Confirmation for the structural assignments was obtained from a series of proton magnetic resonance and mass spectral studies. Application of the new photochemical reaction was suggested to preparation of certain aldehydes otherwise accessible only with difficulty.Chemical degradation of lanosterol to both 14a-methyl progesterone and 14a-methyl testosterone was previously accomplished, as part of an investigation concerned with evaluating the biological importance of 14a-methyl steroids (I). Reconstituting the triterpene-type ring A of lanosterol to one typical of the steroid hormones requires, a t present, a somewhat lengthy series of synthetic steps. With eventual view to providing a greater selection of methods for reconstructing or otherwise transforming the 3P-hydroxy-4,4-dimethyl system of lanosterol, we decided to study photolytic decomposition5 of 30-nitrito5a-lanostane (Ib) .Preparation of 3P-hydroxy-5a-lanostane (Ia) was performed essentially as previously reported (lc), and the required nitrite ester (Ib) was prepared in good yield using nitrosyl chloride (5a) in pyridine. Although the ester proved relatively unstable (5b, 5c) rapid chron~atographic purification on a Florisil column was found suitable for preparing a pure specimen. Irradiation (high-pressure mercury vapor arc lamp) of nitrite Ib in benzene solution was performed under conditions commonly employed for the Barton reaction.5 After approximately 40 min a t 15O, spot tests (diphenylamine in concentrated hydrochloric acid) for the nitrite ester became negative. When a total of 45 min had elapsed, irradiation was terminated and the crude product was isolated. A thin-layer chromatogram of the product indicated that a t least six substances were present. Partial separation of the viscous yellow oil on a column of basic alumina led to two readily predictable products in small yield: the substance expected from alkoxy radical decomposition with a-hydrogen fission (3-0x0-5a-lanostane, 11) and the substance resulting from intermolecular hydrogen abstraction (alcohol Ia). The ketone (11) was finally separated from an oily impurity (later assumed to be nitrile 111) by preparative thin-layer chromatography. However, the 'For Part X X I V refer to: G. R . Pettit, A . K . Das Gz~pta, H. Klinger, and J. Occolowitz. Experientia, 80, 646 (1964). Contribution X X V was abstracted in part from the Ph.D. dissertatiotz submitted by R . E . Kadunce to the Gradzlate School, University of iMaine, M a y 1964. 2Present address: Department of Chentistry, Arizona State University, Tempe, Arizona. 3 N D E A Predoctoral Fellow, 1960