> Hi All, > Im pretty sure its Oxalis eckloniana sonderi. I think that its now In the > O. purpurea group. That is definitely not O eckloniana, and O eckloniana and O purpurea are not closely related. Unfortunately a lavender form of O purpurea has been circulated widely under the O eckloniana sonderii label. The fact that Thunberg originally and in error described O eckloniana var sonderi as O purpurea AFTER O purpurea was described by Linnaes in 1753 may also be a problem, but this has been rectified in 1939 by Salter. (Salter reinstuted the name O purpurea in the place of O variabilis (Jacq), and renamed the O purpurea sensu Thunberg to O exckloniana var sonderii) The two species are easily told apart when in flower - O purpurea have two bracts BELOW the middle of the peduncle (flower stem), while the one or two bracts of O eckloniana are close to the calyx, i.e. rather HIGH on the peduncle. Furthermore, O eckloniana belongs to the section Sagittatae. If you pull the corolla away so that the stamens and styles are left intact, the short and medium stamens or styles opens up and spread wide. Those of the other sections (such as section stictophylla, to which O purpurea (L) belongs) stay close together. The bulbs are also different, O eckloniana usually have smooth light brown tunics, while that of O purpurea are normally very dark to black, gummy and irregular. There are many forms of both species and thus the leaves (number of leaflets and shape), the degree of hairyness and other characters are useless for identification purposes. The Oxalis pictured is probably O. purpurea (L). Christiaan