SYSTEMATIC BOTANY.

481

New Zealand; Camptoloma, from the south of Africa ; Bryodes, from theMauritius ; Synthyris (to which Wulfenia reniformis, Benth. belongs), fromthe west of North America ; Radamiea and Raphispermum, from Mada­ gascar ; Micrargeria, from the East Indies; Synnema = Pedicularis avanaWall., from Ava. Otophylla , Silvia, and Graderia are separated fromGerardia. As regards the special style of Bentham’s Monograph , it ischaracterised by the excellence of the natural division of the genera, andthe appropriate contraction of the forms; the new species are very nu-merous. Webb has published some remarks upon the affinity of the genusCampylanthus Bth., from the Canary Isles (Ann. Sc. nat. 3, iii, p. 33),the position of which Bentham had left doubtful. It differs from theVeronicese in the character of the stamens; in the latter two posterior,and in Campylanthus two anterior stamens are developed, as is also thecase in Anticharis and Achetaria. Webb places them in a distinct group,in which we are more inclined to agree with him than when he wishesto place it near to the Salpiglossidese and Solanacese, from which it differs inthe aestivation.

Solanace^:. Cyphomandra Mart. Solani sp. B. P., described in a mono-graph by Sendtner, differs from Solanum in its large connective (RatisbonFlora, 1845, pp. 161-76). New genera: JochromaBznVa.. = HabrothamnusLindl. ol. (Bot. reg., 1845, tab. xx.), from Ecuador ; Salpichroa and Hebe-cladus Miers = Atropm sp. Amer. austr. (Lond. Journ. of Bot., 1845,p. 321); Lycioplesium and Chcenestes Mrs. = Lycii sp. Amer. austr. (ib.pp. 330-36); Borystigma Mrs. = Jaborosce sp. Chilensis, Hook. ol. (ib.p. 347); Trechoruetes Mrs. (ib. p. 350), from Chili; Pionandra Mrs. =Witheringice sp. Mart., &c. (ib. p. 353).

Nolane^e. In 1844, Bindley, in the ‘ Botanical Register,’ dividedNolana into five natural genera, and characterised the species belonging tothem. Miers has recently also studied the characters of this small group(1. c. pp. 365-469), and described a new type from Chili, Alibrexia (p. 505).Miers regards it as a connecting link between Boraginacese and the Convol-vulaeese : from the former it differs principally in its habit, and in the posi-tion of the embryo; from the latter in its distinct ovaries. Grabowskya (oneof the Boraginacese according to Endlicher, of the Solanacese according toothers) forms the transition to the former, the Dichondrese to the latter. Ifit is desired to preserve a limit between the Boraginacese and the Convol-vulacese, the Nolanese must either be regarded as a distinct family, and theDichondrese placed with them; or, keeping in view the inflorescence and thesestivation of the Boraginacese, both groups must be united with the Convol-vulacese. However, Miers retains Grabowskya as a distinct tribe among theNolanacese, and leaves the Dichondrese with the Convolvulacese.

Ekycibe.®. Both the elder and the younger De Candolle , in the Pro-domus (vol. ix), have separated Erycibe also, another connecting link between

31