The constitution (I) proposed in 1931 by Coyne, Raistrick and Robinson for citrinin, an antibiotic produced byPenicillium citrinum, and the structures assigned to the product (A) of the sulphuric acid hydrolysis of citrinin and the dialkylresorcinol (C) obtained by alkali fusion of (A), have been shown to be untenable in view of their behaviour towards diazonium salts. The dialkylresorcinol (C) is 4-methyl-5-ethylresorcinol, and not 4-methyl-2-ethylresorcinol. From the orientation of the alkyl groups in (C) and the experimental results of Hetherington and Raistrick it follows that Product (A) is 4-methyl-5-(1-methyl-2-hydroxy)-propylresorcinol (IX), as suggested by Cram. Two alternative structures (X and XXV), which are both derived from an isochromane system and which differe from each other only in the position of the carboxyl group, are discussed; the action of diazonium salts on citrinin and on Product (A) is better explained by (XXV).
In the chromatographic adsorption ofmono- andbis-benzeneazoresorcinols on an alumina column, 4-benzeneazoresorcinol is more strongly adsorbed than 2∶4− or 4∶6-bisbenzeneazoresorcinol, while Ruggli and Jensen have observed that the adsorbability of the azo dyes studied by them increased with the number of azo groups. 2-Benzeneazo-4-methylresorcinol is more weakly adsorbed than 6-benzeneazo-4-methylresorcinol. These effects are ascribed to chelation between the azo and hydroxyl groups.