312 PROCEEDINGS OF THE AMERICAN ACADEMY 



pseudo-verticillatis si^atbulatis seu ovato-spathulatis basi longe attenu- 

 atis glabris ; floribus subsolitariis glabris ; filamentis filiformibus sty- 

 loque exsertis. — Feejee Islands ; also collected by Mr. Miliie, and 

 a smaller-leaved form by Prof. Harvey. The leaves resemble those 

 of PepUs Portula. 



Ophiorhiza leptantha (sp. nov.) : fruticosa, fere glabra ; foliis 

 Ifetevirentibus oblongo- seu elongato-lanceolatis utrinque acuminatis 

 longe petiolatis ; stipulis utrinque binis setaceis ; cyma multiflora pu- 

 berula ; floribus plerisque secundis subsessilibus ; corolla alba gracili 

 pollicari, ore tenuissime barbato ; staminibus inclusis ; filamentis an- 

 thera fequilongis ; stylo glabro. — Feejee Islands : also collected by 

 Prof. Harvey. 



Ophiorhiza laxa (sp. nov.) : fruticosa ; ramis junioribus saape 

 ferrugineo-puberulis ; foliis oblongis vel subovatis acuminatis longe 

 petiolatis ; cymis pauci - plurifloris laxis ; floribus pedicellatis ; corolla 

 semipollicari ; cajt. fere praecedentis, sed ramosior laxior. — Feejee 

 Islands : also gathered by Mr. Milne. 



13. Hedyotece. 



Oldenlandia, Houstonia, Hedyotis, etc. It is unnecessary to recapit- 

 ulate the history of the ancient genera Oldenlandia (1703), Houstonia 

 (1737), and Hedyotis (1747), previous to the consummation of their 

 union by Wight and Arnott into one polymorphous, or, as they justly 

 term it, compound genus, for which, following Lamarck, the latest of 

 the three names was adopted. Korthals has since proposed the estab- 

 lishment, or re-establishment, of all Wight and Arnott's principal sec- 

 tions as distinct genera. Bentham, taking an intermediate view, and 

 depending mainly upon dehiscence, adopts two of the Linnasan genera, 

 viz. Hedyotis (to which he refers Dimetia as well as 3facrandia and 

 Diplophragma) and Oldenlandia (to include Houstonia, Anotis, &c.), 

 and also admits Scleromitrion and Kohautia. I had adopted Ben- 

 tham's view, without particular investigation. But I now come to the 

 conclusion, that all three Litmaean genera equally merit restoration, 

 or at least that Houstonia is more definitely distinguishable from Olden- 

 landia than it is from Hedyotis, as the latter is received by Bentham. 

 For the Dimetice and 3fac7-andri(B generally, and many of the Diplo- 

 phragmcB, not only accord with Anotis in habit, in flowers, and in the 

 structure of the seed, but also in the dehiscence, which is loculicidal 



