Notes

Some of the species are very similar in facies to members of the Oldenlandia group of genera but can always be absolutely distinguished by the solitary ovules and discoid pluricolpate pollen grains.It has been accepted practice for a hundred years or so to split up this genus, but although the technical characters of the fruit dehiscence sound admirable on paper there is no associated habit facies and it is not possible without additional knowledge to assign flowering material to the genera concerned. I have therefore followed Hooker in G.P. 2: 145 (1873) in his circumscription of the genus with the addition of Octodon (which he kept separate) to the synomymy as suggested by Hepper (K.B. 14: 260 (1960)). Bremekamp (Rec. Trav. Bot. Néerland. 31: 305 (1934)) who has been responsible for much generic segregation in this family states his position as follows. “The differences between Diodia, Spermacoce and Borreria are also very small and hardly of sufficient importance to justify their separation. In contradistinction with the difference between Diodia and Hemidiodia which is not only taxonomically of little value but also difficult to see the differences between Diodia, Spermacoce and Borreria are at least easily recognizable. For this admittedly purely opportunistic reason I have retained these genera.” I agree entirely but I must report that Steyermark (in litt.), although agreeing with much of this argument, thinks the technical characters are adequate and maintains the genera, thus avoiding a great many name changes since very many S. American species have been described in Borreria. As far as the African species are concerned I would also unite Spermacoce with Diodia and the very close relationship can be seen by comparing Diodia sarmentosa with Spermacoce princeae, but a survey of the New World material of Diodia has convinced me this would not be a wise course—certainly the type species of Diodia, D. virginiana L. differs widely in its fruit structure.I have found no satisfactory logical way to arrange the species, in fact this will be impossible until a thorough revision is made on a world basis. K. Schumann’s classification of Borreria (E. & P. Pf. IV.4: 143 (1891)) as follows, is of no value for arranging the African species.Sect. BorreriaSeries Tenellae K. Schum.Subseries Virides K. Schum. (including Octodon)–S. laevis, S. ocimoides, S. pusilla and S. dibrachiata belong hereSubseries Glaucae K. Schum.Subseries Radiatae K. Schum.—S. radiata and S. chaetocephala belong hereSeries Latifoliae K. Schum. S. latifolia belongs hereSect. Trachiphyllum K. Schum.Sect. Galianthe (Griseb.) K. Schum.