OBOLARIA VIRGINICA. 25 



Statements accords with our observation. The dehiscence of the 

 capsule, however, I have not seen ; but I can scarcely doubt that 

 it is septicidal, or, in other words, that the carpels separate from 

 their margins. Meisner* has followed Endlicher in apjiending 

 Obolaria to the order Scrophulariacese. Grisebach f has neither 

 included it in the order Gentianeae, nor mentioned it among the 

 genera which have been referred to that order; the remark by 

 Nuttall and its adoption by Sprengel having probably been over- 

 looked by him. 



It is manifest, from the foregoing summary of what is on record 

 respecting Obolaria, that its affinities are still unsettled, and that 

 the peculiar structure of the ovary has not been made known. 

 This peculiarity, which I have endeavoured to express in the de- 

 tailed generic character given above, and in the accompanying 

 analyses, was first noticed in the living plant by Professor Torrey 

 and myself, in the spring of the year 1843. 



A view of the transverse section of the ovary, considerably en- 

 larged, is given at Figure 1 1 of the accompanying plate. The 

 parietes of the ovary consist, first, of a thin exterior coat, com- 

 posed of compressed cellular tissue alone, and quite similar to the 

 skin or epidermis, which readily peels from the stem, &c. This 

 coat is but slightly coherent with the parts subjacent, except at 

 the two longitudinal lines, which, alternating with the lobes of the 

 stigma, evidently correspond to the margins of the carpels, and 

 doubtless with the lines of dehiscence at maturity. The outer coat 

 does not follow the introflexions of the interior, or endocarpic, por- 



* PlantcB Vascular es, p. 313. 



+ Genera et Species Geniianearum, etc., 1839, and Gentianacejs in DC. 

 Prodr., Vol. IX. 



4 



