To evaluate growth habits, fresh pod yield potential, and possibility of early and late seeding, seeding dates were extended from March 21 to June 20 by PE mulching and non-mulching. Soil temperature, under 5cm from surface, above 15~circC at 10 a.m. in early seeding reached about March 25 in mulching and April 5 to April 12 in non-mulching. Days to emergence and first flowering were accelerated owing to increasing temperature, as seeding was delayed. Days to emergence according to seeding dates reduced 21 to 8 day in mulching and 33 to 10 day in non-mulching. Days to flowering were ranged from 51 to 26 day in mulching and from 69 to 32 day in non-mulching and differences between mulching and non-mulching on each seeding date had 18 to 4 days. Early seedings till April 21 had 160-170 flowers per plant for 8 weeks, while late seedings from May 21 increased more speedily with 200 flower for 6 weeks. Harvesting of fresh peanut, at 80 days after first flowering, was possible from Aug. 1 to Oct. 7 (133-108 days to harvest) by mulching and from Aug. 19 to Oct. 12 (151 to 114 days) by non-mulching. Yields between mulching and non-mulching in early seeding until April 21 had more difference, but in late seeding after May 21 was higher and showed insignificance. Pod setting periods by early and late seeding were about 3 weeks equally. In late seeding pod setting were almost concentrated for front 15 days. In spite of difference of fresh pod weight between two seeding times, the distributions of average of seed weight showed nearly same tendency.

Near-infrared reflectance spectroscopy (NIRS) was used to estimate the lipid and protein contents in ground seed samples of perilla (Perilla frutescens Brit.) and peanut (Arachis hypogaea L.). A total of 46 perilla and 80 peanut calibration samples and 23 perilla and 46 pea. nut NIRS validation samples were used for NIRS equation development and validation, respectively. Validation of these NIRS equations showed a range of very low bias (-0.05 to 0.13 %) and standard error of prediction corrected for bias (0.224 to 0.803%) and very high coefficient of determination (R2 ) (0.962 to 0.985). It was concluded that NIRS could be adapted as a mass screening method for lipid and protein contents in perilla and peanut seed.