Jim, I agree that there are plenty of intermediate or indeterminate structures that render a rigid classification system impractical. However, if we say that these character states and their terms are mere variations on a single theme, they lose their predictiveness and hence much of their usefulness. What I meant by "replaced" in the case of corms is this: the structure at the start of the growing season, a corm of Gladiolus equitans or a "tuber" of Amorphophallus koratensis for instance, is completely used up in one season of growth. The new or "replacement" corm is derived from *entirely new tissue* formed by the nodes and internodes of the new shoot; the starting corm is only a husk at the end of the growing season. So, the growth is modular. This is a radically different ontogeny than a true bulb or a tuber. As you note, there are plenty of cases where old living corms persist, so the replacement conecpt is not water-tight. But it does apply to a great many taxa, some of which are well-known garden plants. The categorization of these labels is probably of little interest to most people. But what is I think fascinating are hypotheses we can draw up from looking at a range of rootstock forms within different groups. Did cormous plants evolve from rhizomatous ancestors? There seems to be good circumstantial evidence for this in the irids and aroids considering the "intermediate" or even clinal forms found in these diverse families. How did the different types of bulbs arise? One would think the story of bulbous Oxalis-- thank you for that illuminating photo-- is something quite different from the evolution of bulbs in Amaryllidaceae. I prefer to think of the value in recognizing these concepts, and sharpening their delimitation where possible, rather than thinking of them as variations on a single theme. We know too little about the origins of these plants and their rootstocks to say they are all derived from one common process appearing independently in the various groups. Dylan