The paper by Kramer (1999) on circular xylem (whirled grain) as an example of topological defects warrants several comments. (1) Whirled grain is not disordered as claimed, but rather a di!erent order compared with straight wood. Moreover, such grain occurs regularly in many branch junctions: in small annual plants such as the model plant Arabidopsis thaliana (Lev-Yadun, 1996) and in many tree species (Lev-Yadun & Aloni, 1990) (Fig. 1). (2) The cambium is not a partially ordered system, but rather a fully ordered one, with the ability to change in response to environmental (biotic and abiotic) cues, as well as to age and organ-related internal variations (Larson, 1994). Two types of initial cells occur in the cambium: fusiform initials, which divide and di!erentiate to form the axial system, and ray initials, from which the radial system (the rays) di!erentiates (Philipson et al., 1971;Larson, 1994). Fusiform cambial initials can reduce through cell division into ray initials, and ray cambial initials can elongate into fusiform initials (Philipson et al., 1971). The ratio of fusiform to ray cells in the cambium varies greatly between species, and within the same plant at di!erent locations, and following wounding. The phytohormone ethylene seems to be the major regulating factor in the conversion of fusiform cambial initials into ray initials (Lev-Yadun & Aloni, 1995). The structure of the secondary