Fig. 1. The inhibition of hydrogen bonding between the hydroxyl group in the 5-position and the 4-carboriyl group gives azaleatin, the property possessed by flavonoids lacking a 5-hydroxyl group (that is, fisetin and robinetin) of intense fluorescence in the ultra-violet. Thus, azaleatin is readily detected in hydrolysed plant extracts by paper chromatography, and although its RF (0-49) is close to that of quercetin (0-42) in the 'Forestal1 solvent, it is clearly distinguishable from quercetin and other common flavo-nols in other solvents (Rp&in butanol-acetic acid-water are 0-48 and 0-64, respectively). The records obtained during surveys of the phenolic constituents in the leaves of dicotyledonous5 and monocotyledonous plants (unpublished results of Bate-Smith) showed that a substance with this unusual fluorescence at the appropriate RF occurred in only one out of more than a thousand plants examined, that is, in Eucryphia glutinosa (Poepp. et Endl.) Baill. Detailed examination of the pheriolics of this plant has confirmed the presence of azaleatin in the leaf. The isolated aglycone agreed in every respect with authentic material. Azaleatin, known in Rhododendron and Plumbago as the 3-rhamnoside, azalein1'2, is present in Eucryphia as the 3-galactoside and as a 3-bioside, probably the 3-arabinosylgalactoside, and is accompanied by the 3-galactoside of quercetin. A third constituent in E. glutinosa (Rp 0-80 in 'Forestal') had the blue colour in ultra-violet normally associated with caffeic esters and related compounds, but spectral examination (Xmax 253 and 347 mjji) showed it to be a quercetin derivative. Its spectral properties indicated that it was the 3,5-dimethyl ether, for it gave a positive sodium acetate shift (indicating a free 7-hydroxyl grouping), a positive boric acid shift (indicating a free 3',4'~ dihydroxyl) and a stable alkaline shift (3-hydroxyl blocked), but it gave no aluminium chloride shift, showing that the 3- and 5-hydroxyls were substituted. Its identity as the 3,5-dimethyl ether, caryatin, was confirmed by demethylating it to the 3- and 5-monomethyl ethers and eventually to free quercetin.Azaleatin and caryatin were also found in a herbarium specimen of Eucryphia cordifolia Cav. and in the fresh leaves of E. nymanensis (E. cordifolia x E. glutinosa} and E. x intermedia Bausch (E. glutinosa x E. lucida}, but they were absent from E. lucida (Labill.) Baill. itself, from E. moorei F. von Mueller and from E. milliganii Hook f. The latter species had a constituent agreeing in all respects with dihydroquercetin, 3,5,7,3',4'-penta~ hydroxyflavanone (Rp in 'Forestal' 0-77, dark in ultraviolet becoming mustard yellow when fumed with ammonia and turning red in visible light after prolonged fuming). The three species lacking azaleatin and caryatin are from south-eastern Australia and Tasmania, whereas the other two species are South American, so that there is an apparent correlation between the occurrence of quercetin methyl ethers and plant geography in this genus; the above survey represents a complete coverage of the genus as at present known. The discovery of azaleatin in the Eucryphiaceae brings the number of families with this character to four; the others are the Juglandaceae, Plumbaginaceae and Ericaoeae. These families are similar in being predominantly woody in character, and although they are placed far -apart in most systems they all show closer affinities with the Rosalean families (for instance the Leguminosae) than their present positions suggest.