“…The aqueous layer was further extracted with 1-butanol (1-BuOH) to give a 1-BuOH (12.5%) and a H 2 O (11.0%) soluble fraction. The EtOAc-and 1-BuOH-soluble fractions were subjected to normal-phase and reversed-phase silica gel column chromatography and repeated HPLC to give florahydrosides I (1, 0.068%) and II (2, 0.014%), thunberginol (3R)-thunberginol I 4′-O-β-d-glucopyranoside (10, 0.056%), 18) (+)-3-(4-methoxyphenyl)-8-hydroxy-3,4-dihydroisocoumarin (11, hydrangenol monomethyl ether, 0.011%), 19) phyllodulcin (12, 0.19%), 20,21) hydrangenol 8-O-β-d-glucopyranoside (13, 2.87%), 16,18) thunberginol G 3′-O-β-d-glucopyranoside (14, 0.16%), 16,22) hydrangenol (15, 4.79%), 23,24) thunberginol G (16, 0.067%), 5,25) neochlorogenic acid (17, 0.075%), 26,27) 29) chlorogenic acid (20, 0.054%), 30) chlorogenic acid methyl ester (21, 0.049%), 9,31) taxiphyllin (22, 0.076%), 8,32,33) umbelliferone glucoside (23, 0.11%), 34,35) α-morroniside (24, 0.021%), 36) trans-p-coumaric acid (0.013%), 30,37) and shikimic acid (0.36%) 38) (Fig. 1 C-NMR spectra of 1 were superimposable on those of (3R)-thunberginol E [(3R)-3,4-dihydro-6,8-dihydroxy-3-(3-hydroxy-4-methoxyphenyl) isocoumarin], 22) except for the signals around the 8-position, while the proton and carbon signals due to the dihydroisocoumarin part including glycoside moiety were very similar to those of…”