Page:A Voyage to Terra Australis Volume 2.djvu/573

Natural Orders.] the primary branches; but in other cases where they are equally distinct at the base of the tube, this supposition cannot be admitted. A monopetalous corolla not splitting at the base is necessarily connected with this structure, which seems also peculiarly well adapted to the dense inflorescence of Compositæ; the vessels of the corolla and stamina being united, and so disposed as to be least liable to suffer by pressure.

As this disposition of vessels is found in Ambrosia and Xanthium they ought not to be separated from Compositæ as Richard has proposed; and as it does not exist in Brunonia I prefer annexing that genus to Goodenoviæ, with which it agrees in the peculiar indusium of the Stigma.

GOODENOVIÆ. This order I have formerly separated from Campanulaceæ, considering the peculiar membranous cup surrounding the stigma, along with a certain irregularity in the corolla, as sufficient distinguishing characters, especially as these are accompanied by other differences which appear to me important. In Goodenoviæ I have not included Lobelia, which, however, has also an irregular corolla, and although it wants the peculiar indusium of the stigma, has in its place a fasciculus or pencil of hairs surrounding that organ. This structure has been regarded by Jussieu and Richard, in a very leared memoir, more recently written on the subject analogous to the indusium of Goodenoviæ, to which they have therefore added Lobelia and derived the name of the order from this, its most extensive and best known genus. To the opinion of these authors I hesitate to accede, chiefly for the following reasons:

1st. In Goodenoviæ the deeper fissure of the tube of the corolla exists on its inner or upper side, a circumstance readily determined in those species having simple spikes. In Lobelia, on the other hand, the corresponding fissure is on the outer or lower side, a fact, however, which can only be ascertained before the opening of the corolla, the flowers in the greater number of species becoming resupinate in the expanded state, a circumstance that does not appear to have been before remarked. The relation therefore not only of the corolla but of the calyx and stamina to the axis of inflorescence, is different in these two tribes.