Page:Miscellaneousbot01brow.djvu/404

386 Nepenthes, which has a superior ovarium, is related to Asarinæ, as I am inclined to believe; and on the other, by considering Homalinæ, whose ovarium is inferior, as allied to Passifloreæ, the order with which I shall now proceed to compare Rafflesia.

The comparison is suggested by the obvious resemblance between the perianthium of our genus, and that of certain species of Passiflora itself; or of other genera of the order, as Deidamia, in which the inner series of segments is wanting. Thus, they agree essentially, and even remarkably, in æstivation of perianthium: the corona of Rafflesia may be compared with that of Murucuia, and the two annular elevations at the base of the column with the processes of like origin and nearly similar form in some species of Passiflora. The affinity is also supported by the position of the stamina on a central column.

The peculiar structure of antheræ in Rafflesia can hardly be regarded as an objection of much weight to the proposed association; and it will at least almost equally apply to any other family with which this genus may be compared.

If the concentric processes on the disk of the column in our plant are to be regarded as indications of the number and disposition of pistilla, or of the internal structure of ovarium in the female flower, they present a formidable objection to its affinity with Passifloreæ, in all of which the ovarium is unilocular with parietal placentæ. If, however, these processes were considered as inner series of imperfect stamina, the objection derived from their number and arrangement merely, would be comparatively slight; for in some genera of Passifloreæ, particularly in Smeathmannia, the stamina are also numerous and perhaps even indefinite.