Relatively few studies have been made of changes in the flavonoid constitution of plants as they develop. Bate-Smith (2) has presented a qualitative picture of such changes in the maturing flower of Coreopsis. Hesperidin (5) and naringin (8) in citrus fruits and rutin in Nicotiana (1) have been examined quantitatively under field conditions.Since rutin has become of commercial interest, a number of workers have followed the formation of this glvcoside in mature buckwheat plants (7,10,11,13 ). Workers of the United States Department of Agrictulture found that in Japanese (3,12) and Tartary (4,12) The plants were then grouped into four lots, two of 50 to 100 and two of 200 to 300 individuals. As rapidly as possible all were divided into their constituent organs. As soon as all free water appeared to be gone from the plant surfaces, fresh weights were obtained for all four lots. The two smaller groups were then dried at 680 C for 48 hours and stored in a desiccator until dry weights could be measured. The two larger lots, to be used for the flavonoid analyses, were placed immediately in 95 % ethanol and kept in this solvent. Such a procedure was followed to eliminate possible partial destruction of the flavonoid substances during drying (12).EXTRACTION: The plant material stored in 95 % ethanol was extracted (Soxhlet) with additional portions of this solvent. The extracts of each plant part for each harvest date were combined, concentrated at reduced pressure, and made to a convenient volume (5 to 50 ml).
AN-ALYSIS:The analytical method used was that previously described (15). Although it should be regarded as semi-quantitative in nature, it was the only procedure which the author was able to apply with success to the mixture of flavonoids present in buckwheat seedlings. Aliquots of the extracts were spotted on strips of Whatman No. 1 filter paper 2.5 cm by 57.5 cm; these were equilibrated in a glass chromatographic chamber for 24 hours and then developed (descending) with the upper layer of the system chloroform-isobutanol-water (2-4-4 by volume). After drying, the strips were scanned at 350 mu in a Beckman DU spectrophotometer. For each sample three replicate strips were run on different days, routine procedures being randomizel. Quantities of rutin w-ere calculated from observed areas under plots of absorbance vs. distance on chromatogram and from the previously obtained standard curve for rutin.