Cl. 11.] RUBTACE^. 127 



Brown observes, Bot. of Terra Australia, 3 1 , that it 

 is scarcely possible to distinguish the Rubiacete, as 

 now constituted, from the Apodnece, Ord. 47, by cha- 

 racters taken from the fructification alone. This is 

 but one confirmation amongst many, which the nu- 

 merous exceptions throughout the Jussieuan classifi- 

 cation afford, of the opinion of Linnaeus, that natural 

 orders are, as yet, not possibly to be defined by tech- 

 nical marks. Nevertheless, every attempt of the 

 kind is useful, as tending to dissipate some obscurity, 

 or to point out some truth ; nor does the fact just 

 mentioned at all invalidate the propriety, or necessity, 

 of recurring to the fructification, for every principle of 

 classical arrangement, as well as of generic distinction, 

 though our incomplete knowledge of plants renders 

 exceptions, to all our rules, inevitable. 



Sect. 1 . Fruit of 2 single-seeded grains. Stamens 

 mostly four. Leaves mostly whorled, and Stem her- 

 baceous. Sherardia, fig. 196, Aspcrula, Galium, Gru- 

 cianella, Valantia, Rubia, and Anthospermum, the 

 last not well characterized by Linnasus. 



Sect. 2. Fruit the same. Stamens 4, rarely 5 or 

 6. Leaves generally opposite, connected by a fringed 

 sheath. Stem usually herbaceous. Houstonia, Knoxia, 

 Sperwacoce, Diodia, Galopina Thunb. Rickardia, and 

 Phyllis. 



Sect. 3. Pericarp simple, of 2 cells, with many 

 Seeds. Stamens 4. Leaves opposite. Stem herba- 

 ceous or shrubby. Hedyotis and Olcknlandia, two 



