The shape of the flower on M. longiflora is quite distinct from that of M. jalapa, which is the only other species with which I'm familiar. Those of M. longiflora are daintier in appearance, small in relation to the length of the perianth tube which can be 2 inches or more in length. Also, the flowers come in groups of three, while they appear individually in M. jalapa. My plant has, as I said, white flowers with attractive thread-like violet stamens. The only alternate flower color given in my reference books is "tinged or marked with pink or violet." No yellow. If you knew how many cannas I was digging each autumn (boxes and boxes of them) you'd understand that one mirabilis was not a big deal (except when you chop part of it off.) My plant makes those large dark brown seeds, clasped in a sticky calyx. They often drop to the ground but I've never seen self-sown seedlings. Which could mean it is forming seeds but perhaps they are not viable. I've never bothered sowing seed as one large pot, and storing it over the winter is sufficient. My plant has not been blooming, which I attribute to the very overcast, rainy weather we've been having. The pot is under the roof overhang, so it is not getting soaked. If it does bloom and seed is set I'll sent some to Dell. Does anyone think it would propagate from stem cuttings? If you saw the rootstock / tubers, you'd understand that division is not an option. Alberto, compact is just about the antithesis of how I'd describe M. longiflora, which is a sprawler and a flopper. M. dichotoma, from Mexico, is said to have yellow flowers but not much more is listed about it. And there's M. X hybrida, a cross between M. jalapa X M. longiflora listed in the RHS Dictionary, with "Infl. dense, fls white, sometimes marked or tinted crimson or yellow." Judy