48 Dansk Botanisk Arkiv, Bd. 2. Nr. 8. 



Bentham and Diels) show great uncertainty with regard to the 

 delimitation of the species. Niedenzu, in his survey of the 

 Frankeniacæ in Engler u. Prantl., Natürl. Pflanzenf. (Ill, 6, 1895, 

 pp. 283 — 289) divides the subgenus Oceania, to which all the 

 Australian species belong, into two sections, according to the 

 number and place of the ovules. After examination of a number 

 of specimens of West Australian Frankenias in the herbaria of 

 Copenhagen and Berlin, I have found this distinction mark a 

 good one. 1 I distinguish between: (1) ovules on 2—3 parietal 

 placentas, more than two on each placenta; (2) ovules 1 — 2 on 

 each parietal placenta; and (3) ovules only 2—3, basal on long 

 funicles. These three groups fit in for the West Australian species, 

 but I have examined a plant from Port Pirie, S. Austr., which 

 seemed to have 9 basal ovules. My two first groups are Niedenzu's 

 Toichogonia, and my last group his Basigonia. 



Using these characters as the principal ones, and then those 

 given by Bentham in his very good treatment of the genus in 

 Fl. Austral, vol. I, I have arrived at the conclusions given in the 

 following. 



I have found it necessary to describe some new species, and 

 I feel convinced that further researches will result in still more 

 new discoveries. As Diels, in his Fragm. Phytogr. Austr. occ, 

 remarks, F. pauciflora belongs to "einen polymorphen Formen- 

 Kreis", and my examination of specimens labelled F. pauciflora 

 have shown that some belonged to sect. Toichogonia and others 

 to Basigonia; these must of course be separated from each other. 



It is not possible yet to arrange the species of Frankenia 

 into natural groups, as our knowledge is too restricted, and the 

 following key to the West Australian species must, therefore, be 

 taken only as an arrangement according to the characters most 

 easily used for distinction. 



A Key to the West-Australian Frankenias. 

 A. Placentas parietal, each bearing several (more than 2) ovules. Stems 

 decumbent to erect; flowers large or smaller. 



a. Leaves shortly, but distinctly petiolate; flowers in leafy dichotomous 

 cymes. 



1. Leaves linear, with margins revolute until the midrib; flowers 

 large; stem decumbent or ascending, glabrous or pubescent; calyx 

 glabrous or pubescent F. pauciflora. 



2. Leaves, at least the lower ones, ovate with only the margins re- 



1 Niedenzu, by the way, not having sufficient material at his disposi- 

 tion, has arranged the Australian species wrongly under his headings. 



