Nhu mentioned Triteleia and Brodiaea as genera where seed ripens after the inflorescence is detached. I notice this particularly with the related Dichelostemma, which species have long scapes (main stems). By the time the seed is partially ripe, the scape is detached from the corm, and the moisture in the above-ground parts seems to be sufficient to maintain the plant's energy until the seed is ripe. I wonder if this is an adaptation to avoiding browsers. If the stem and seedhead are lying flat on the ground, deer would be less likely to eat them -- though rodents would be more likely to carry off the seeds, which might then have the opportunity of germinating in an abandoned rodent cache. Although few composites are within our purview, I'd note that it's possible to harvest what appears to be half-ripe seed from many of them, particularly alpine species, and it will ripen to the point of viability later. I have collected both Rhodophiala and Alstroemeria capsules that were still greenish, and they ripened in the paper envelopes (don't put them in plastic). In fact, you should get the Alstro capsules a bit early, because when fully ripe they explode. You can hear them popping in their envelopes, and also rosulate Viola will do this, making you think you've got a mouse in your room. (Well, considering the hotels one is likely to be in after collecting these seeds, you probably do have a mouse, if not something worse.) If you grow Kirengeshoma and want to propagate it from seed, you have to plant the seed when it appears to be far from ripe, in late fall, when it's soft and moist. Understanding the life cycles and climate regimes of different plants gives us many clues about how to handle their seeds. Jane McGary Portland, Oregon, USA