34 AN ENUMERATION OF THE BRUNIACE^ [Louchostoma 



3. L. ACUTiFLORUM Wickst. op.cit. 353, t. 10; DC. op.cit. 580; 

 Sender, op. cit. 317. Ptyxostoma myrtoides Vahl, op. cit. 97. 



Coast Eegion. — Tulbagh Div. : Mostert's Hoek Mountain, 

 near Mitchell's Pass, 2700 ft., Guthrie, Herb. Bolus and Keiv, 

 3233! 



4. L. QUADRiFiDUM O. Kuntze, op. cit. 86 ; K. Schum. in Just, 

 Jahresb. xxvi. i. 343 (1900). Ptyxostoma quadrifidum Kuntze, I. c. 



Coast Eegion. — Caledon Div. : Caledon, Kuntze, Herb. 

 Berlin. 



The species comprising the genus Lonchostoma depart widely 

 from the general structural peculiarities characteristic of the 

 Bruniaceae in their gamopetalous corollas and epipetalous stamens, 

 but despite this they have been allocated to the choripetalous 

 members of the Order by general consent, nor is anatomical 

 investigation opposed to their inclusion in this group." Their 

 gamopetaly is explained by various writers as due to the fusion of 

 the staminal filaments with the claws of the petals; a not unlikely 

 hypothesis in view of the alternation of the stamens and petals 

 within the Order. 



Endlicher and A. De Candolle assumed an affinity of this group 

 with the monotypic Betzia, and accordingly placed it within the 

 tribe Eetzieae of Solanaceae, but this view is untenable and incon- 

 sonant with the oligospermous nature of the two-celled ovary ; 

 moreover, anatomists are fully agreed that Betzia is misplaced, 

 and shows but little affinity with the Solanaceae, and after enjoying 

 an isolated systematic position in various groups, has at last been 

 placed by Fedde t near Nuxia among the Buddleioideae of Logani- 

 acese, a position undisputed by subsequent writers. 



Considerable variation is displayed by the species in the 

 number of their ovules ; I have invariably found two in each cell 

 of L. obtiisiflorum Wickst., eight in L. acutifloruni Wickst., while 

 in L. monostylis Sond. the number fluctuates from six to ten. In 

 its subsessile, hairy anthers and monostylar character the latter 

 species deviates considerably from its allies, which led E. Meyer 

 to propose the names Peliotis or Peliotus, but it appears inex- 

 pedient to revive these, inasmuch as in its other characters it 

 conforms to the other members of the group, while the single 

 style undoubtedly illustrates a case of fusion, which is evinced by 

 the longitudinal furrows ; the same character obtains in Berzelia 

 and Staavia, and may be similarly interpreted from 2-styled 

 forms, such as Brunia and Pseudobcechea respectively. 



Niedenzu j states that the anthers are hairy throughout the 

 genus, but I have been only able to verify this for L. monostylis 

 Sond. 0. Kuntze describes the coralline- throat of L. quadrifidum 

 as hairy. This species is particularly interesting as exhibiting a 

 partial case of tetramery, partial in so far as the corolla and 

 androecium are concerned, while the calyx is 5-fid. Complete 



* Solereder, Syst. Anat. Dicotyl. 658, 1899. 



t Fedde, Vergl. Anat. Solan. Diss. Breslau, 1896. 



I Op. cit. p. 134. 



