Page:Miscellaneousbot01brow.djvu/578

 560 ON THE RELATIVE POSITION OF

arrangements the adaptation to the performance of function is equally manifest.

If the correctness of these observations be admitted, it follows that characters dependent on the various mocUHca- tions of stigmata are of less value, both in a systematic point of view as determining the limits of families, and theoreti- cally in ascertaining the true composition of organs, than those derived from the analogous differences in the ovaria or placentae.

In those cases in which the nature of the composition of the ovarium is doubtful, it may, in the first place, be re- marked, that wherever in the compound unilocular pistillum the placentae are double or two-lobed, it is more probable that such placentae are derived from two adjoining carpels, and are consequently marginal or submarginal, than that they occupy the disc of one and the same carpel ; this being entirely the appearance in many cases where the marginal origin of placentas is admitted ; while in the greater part of those in which the disc is known to be ovuliferous, the ovula are never collected in two distinct masses, being generally scattered equally over the surface.

But the double placentae are manifest in Orchidece, the principal family in which ]Mr. Lindley considers the ovula as occupying the disc and not the margins. In this family also the alternation of stigmata with placentae is that rela- tion which is most usual in compound unilocular ovaria, where the apparent number of stigmata and ])lacentae is equal ; and that in Orclddece each apparent stigma is formed bv the confluence of the two stionnata of one and the same carpel, is proved by tracing to their origins their vascular cords, Avhich are found to coalesce with those of the three outer foliola of the perianthium.

This view of the composition of the ovarium in Orcltldcce is confirmed by finding that it agrees with the ordinary arrangement in monocotyledonous plants ; namely, the opposition of the double parietal placentae to the three inner divisions of perianthium^ while in Aposfasia the three pla- centae of the trilocular ovarium are opposite to the three outer

1 Deuham, Trav. in Afr. Append, p. 243. {AnU, p. 300.)

�� �