24 CHLORIS BOREALI-AMERICANA. 



above cited, has given a tolerable figure of the plant, follows Nut- 

 tall in referring the genus to Gentianese. Sprengel appears to be 

 the only succeeding author who has adopted this view.* Elliott, 

 although he has introduced the genus under the class Didjnamia, 

 states that the plant, " from the structure of the corolla and the in- 

 sertion of the stamens, certainly belongs to the class Tetrandria." t 

 He makes no remark respecting its natural affinity. But in his 

 account we meet with the earliest, and indeed the only, indica- 

 tion of any peculiarity in the structure of the ovary. He describes 

 the capsule as " four-celled ? or perhaps one-celled with the rudi- 

 ments of partitions." 



The late Professor Don, J in a revision of the order Orobanchese, 

 appends to it a tribe Obolarice, comprising Obolaria and the (totally 

 unlike) genus Tozzia, which are merely said to differ from Oro- 

 banchese proper in being terrestrial instead of parasitical. Bart- 

 ling § also enumerates the genus Obolaria under Orobanchese. So, 

 likewise, does Lindley, both in his Introduction to the Natural Sys- 

 tem, and in the recent Vegetable Kingdom. Endlicher, on the other 

 hand, has placed the genus among the " Scrophularineis affinia," 

 remarking that it appears not to belong to the Orobanchese, but 

 may perhaps be referable to the Gentiane3e.|| The structure of 

 the ovary and of the capsule are described by Endlicher in some- 

 what incongruous terms. The ovary is said to be one-celled, ivith 

 two parietal placentfB, while the capsule is said to be two-celled, two- 

 valved, the valves placentiferous in the middle. Neither of these 



* Genera Planiarum, Vol. I., p. 110. 



+ Sketch of the Botany of South Carolina and Georgia, Vol. II., p. 134. 



t In Edinb. Phil Journ., Vol. XIX., p. 113. 



§ Ordines Naturales, p. 174. 



II Genera Planiarum, p. 695. 



