( 164 ) 



Observations on the Affinities of Vellosia, Barbacenia, GlauiC, 

 Aucuba, Viviania, Deiitzia, and of a New Genus of the or- 

 der RubiacecB. By Mr David Don, Librarian to the Lin- 

 nean Society, Member of the Imperial Academy Naturse 

 Curiosorum, of the Royal Botanical Society of Ratisbon, and 

 of the Wernerian Society of Edinburgh, &c. (Communicated 

 by the Author). 



VELLOSIA AND BARBACENIA. 

 These genera, together with Xerophyta, I consider as form- 

 ing part of the family of HypoxidecB, to which they appear to 

 me to bear a greater affinity than to either of the other families 

 with which they have been associated. They agree with Hy- 

 poxidecB, in having a monophyllous perianthium, whose tube is 

 completely adherent to the ovarium, which is therefore wholly 

 inferior ; in the stamina being inserted in the base of the divi- 

 sions of the perianthium, which are disposed in a double series ; 

 in the structure and insertion of the anthers ; in their trilobate 

 stigma ; in the presence of a fleshy, epigynous disk ; in their 

 trilocular, polyspermous ovarium ; and finally in the seeds being 

 furnished with a prominent umbilicus. Their habit also cor- 

 responds better with HypoxidecB than with HcBmodoracecB, in 

 which the inflorescence is panicled ; and the ovarium little more 

 than half inferior. Dr Martins, who, in his elegant work on the 

 plants of Brazil *, has referred Vellosia and Barbacenia to the 

 Hcemodoracea, describes their seeds as being furnished with a 

 thin membranous testa ; but, from an examination of the seeds 

 of Vellosia Candida, although not perfectly mature, I am led to 

 beheve that the testa, when examined in the mature seeds, will 

 be found to be crustaceous, like that of the Hypoxideoe. As 

 the consistence of this organ appears to afford the only certain 

 mark of discrimination between some families of the great class 

 of Liliacea, it would be highly interesting to know its structure, 

 in the ripe seeds of Vellosia, Barbacenia, and Xerophyta, as, 

 without such knowledge, whatever opinion may be advanced 

 respecting their affinities, will still be conjectural. I have ascer- 

 tained, however, that the seeds of Vellosia and Barbacenia, are 

 furnished with a very short process, analogous to the rostelliform 



• Nov. Gen. et Sp. PI. Bras. i. p. 13. 



I 



