Page:The Botany of the Antarctic Voyage.djvu/314

282 ''DC. Prodr.'' vol. iv. p. 53. ''Hook. Ic. Plant.'' t. 16. Polycarpon Magellanicum, ''Linn. fil. Suppl.'' p. 115. ''Forst. Comm. Goett.'' vol. ix. p. 23. t. 3. Orites depressa, ''Banks and Sol. in Mus. Banks, cum icone''.

South Chili and Fuegia; from the Chonos Archipelago to Cape Horn; Forster, Banks and Solander, Capt. King, C. Darwin, Esq., J. D. H.

I have examined several specimens of this plant in the collections of Capt. King and Mr. Darwin, and find uniformly three stamens and three styles; the flowers seem occasionally incomplete, and then are scarcely half the size of the fertile ones, and with much shorter stamens and styles. Tube of the calyx conical, its limb irregularly cut into 5-7 segments of unequal length, there are also one or two bracte&aelig;, so placed on the ovarium that they are with difficulty distinguishable from the calycine lobes. The petals are also inserted with the latter and they seem to pass the one into the other; they are white and of unequal size, much smaller in the incomplete than in the fertile flowers. Within the petals is a broad flat disk, from the centre of which arise the stamens, three in number, alternating with, but almost united at their base to, the three long diverging styles: the filaments are elongate and subulate; the anthers extrorse; the pollen globular, rough, and containing an obscurely three-lobed nucleus. Ovarium three-celled, each cell bearing at the upper portion of the inner angle a somewhat fleshy placenta, covered with many series of horizontal or ascending ovules. Apparently only a few seeds ripen in each cell.

Donatia is a very anomalous genus of Saxifrage&aelig;, though I entirely agree with M. St. Hilaire that it does belong to that Order. The gradual passage of the leaves into bracts, of the bracts into calycine lobes, and of these again into the petals, together with the adnate tube of the calyx, and the extrorse anthers, (a sufficiently obvious, but hitherto unnoticed character, foreign to Saxifrage&aelig;), are what may be observed in Calycanthe&aelig;. The analogy between this plant and Stylide&aelig;, through Forstera, is very striking, especially in the tufted habit and linear leaves, in the often abortive flowers, the form of the inferior ovaiy, and the irregular insertion of the calycine lobes and of the bracts on the tube of the calyx, in the variable number of the unequal petals, which, though gamopetalous in Forstera and Stylide&aelig;, are occasionally separate in the allied Orders Lobeliace&aelig; and Goodenovie&aelig;, in the flat disk intervening between the base of the petals and the stamens, in the close application of the filaments to and their alternation with the three styles, altogether forming, as it were, one body in the centre of the flower, in the extrorse anthers and somewhat too in the form of the pollen; in the axile placentation and numerous ascending ovules, and in the form of the seed, so far as I can compare it with immature ones of Forstera clavigera, excepting that the raphe in Donatia is prominent.

The Donatia fascicularis is very abundant throughout Fuegia and the western portions of South Chili, covering the surface of the ground in densely-matted and widely-extended, hard, bright-green patches: composing, with a few other plants, such as Caltha appendiculata and particularly Astelia pumila, the greater proportion of the peat-earth in those countries. It is one of the few bog plants, characteristic of the South Fuegian Flora, that has not been detected in the Falkland Islands.

1. AZORELLA, Gaud.

&sect; I. Caulibus dense fastigiatis, foliis subsessilibus arcte imbricatis coriaceis, umbellis brevissime pedunculatis:&mdash;Chamitis, Banks et Sol.

1. c&aelig;spitosa, Cav.; dense c&aelig;spitosa, foliis arete imbricatis basi latioribus vaginantibus patentibus subrecurvis linearibus acutis integerrimis coriaceis vagina filamentosa, umbella florente abbreviata inter folia summa sessili, calycis tubo piloso margine obtuse et brevissime 5-lobo, fructu immaturo oblongo subquadrato dorso compresso. Azorella c&aelig;spitosa, ''Cav. Ic.'' vol. v. p. 57. t. 484. f. 2. ''Poiret, Encycl. Suppl.'' vol. i. p. 551. ''DC. Prodr.'' vol. iv. p. 77. A. gummifera, Poiret, 1. c. A. crassifolia, Persoon, Syn. vol. i.