This page needs to be proofread.

292 OBSERVATIONS ON THE NATURAL FAMILY

1. Sal men sea n dens, (Decand. 1. c.) in which the aristae are equal and without any membranaceous border : stig- mata remarkably dilated, tongue-shaped, obtuse, not hispid, obscurely papulose, and apparently without any terminal appendix : style dilated at the base into a hemispherical bulb which is truncated underneath.

2. Salmea hirsufa, (Decand. 1. c.) whose aristae are un- equal ; the inner, which is the larger, being furnished with an evident ala ; the outer having a narrow margin only : stigmata sharp and spreading : style dilated into an ovate bulb which has an attenuated base.

3. Salmea? curviflora (nob.) differs from both the pre- ceding in the tube of its corolla being remarkably bent outwards. In place of the inner arista there is a broad obtuse wing, of which the inner margin is straight and thickened, the outer continued down nearly to the base of us] the pericarpium ; the outer arista is winged : and besides these, one or two minute processes are generally observable. Stigmata revolute. 1

1 In the remarkable character of its re-curved florets, as well as in some other respects, this species of Salmea agrees with Spilanthus arboreus of George Forster (in Commentat. Gotting. ix, p. 66), of which he originally formed his genus Laxmunnia ; from a very erroneous view of its structure, however, having described the Nectarium or glandula epigyna as a "germen superum ;" the real, though imperfect, germen with its two aristae as a " perianthium bidentatum," and consequently referring the genus to Polygamia segregata.

When he afterwards corrected these errors, and reduced Laxmannia to Spi- lanthus, he did not discover that he had only the imperfect hermaphrodite or male plaut before him.

That Spilanthus arboreus is really dioecious, I have ascertained from the exa- mination of numerous specimens collected by Sir Joseph Banks in the Island of St. Helena, where it forms a small tree called by the inhabitants "White-wood. It is Bidens arborea and perhaps also Spilanthus tetrandrus of Dr. Roxburgh's List of Plants appended to General Beatson's Tracts on St. Helena ; the former being probably the female, the latter a starved variety of the male plant.

In re-establishing Spilanthus arboreus as a genus, sufficiently distinct from Bidens, Spilanthus, and Salmea, it will not, I conclude, be considered expedient to recur to Porster's name Laxmannia, which as far as relates to this plant is connected only with a series of blunders, was abandoned by the author himself, and has since been applied to another genus already generally adopted. It may be distinguished by the following character, and named

Petrobium.

Involucrum polyphyllum subduplici serie : exteriore breviore, foliolis pau- cioribus. Receptaculum paleaceum, planiusculum. Flosculi dioici, tubulosi,

�� �