Dear Sir:

I take the liberty of sending you a short article on Oxalis violacea, which may interest you.¹ Since it was published I have received a note from Prof. J. E. Todd² of Beloit, Wis. (perhaps 100 miles from Madison), stating that in an examination of 25–30 plants in that region, last spring, he found but the two forms described.

Perhaps one of the most convincing facts is that in table III the pollen grains from the two sets of stamens differ in average diameter less than 1.5 μ, the corresponding difference in table IV being but 1.3 μ; while the grains from the long stamens of III and the short stamens of IV (which should correspond, approximately, in a trimorphic species) differ by about 5.45 μ, In case the species really is trimorphic, the long stamens of the long-styled flower, and the short stamens of the short-styled flower—have become displaced considerably, so that the mid-styled form (if ever found) should have an intermediate length of styles, not far from 3.4 mm.³ As it is, the two forms are so constructed that either length of pistils is fertilized by pollen from two sets of stamens one of which is slightly longer the other slightly shorter than itself. This suggests to me that the dimorphism has been derived from a previous trimorphic state, by compensating modifications of the long stamens of the long-styled form, and the short stamens of the short-styled form, and their pollen—, after the mid-styled form had for some reason become extinct.

When studying the flowers I was struck by the differently-placed nectar found in the two forms.⁴ As yet I am not fully convinced whether the nectar is secreted by the petals or stamens; both organs are said to be nectariferous in Oxalis, by different writers, and I was unable to satisfy myself as to the species in question.

Very respectully, | Wm. Trelease.