Page:The Botany of the Antarctic Voyage.djvu/266

234 long and consists of a single stem (probably one of several from the same root) bearing linear, obtuse, and entire radical leaves, scarcely an inch long ; the ascending portion is erect and scarcely branched, with three sessile, ovato-oblong leaves ; the raceme 3 inches long ; pods erect, 4 lines long and less than one in breadth, borne upon stout pedicels shorter than themselves ; seeds immature, small, pale red brown. In this genus, characters taken from a solitary though complete specimen are invalid, and since it is undistinguishable from some North American and European states of B. incana, I am obliged to unite it with them ; for I can find no reason to suppose that they will eventually prove distinct.

It is very singular that this plant, apparently identical with one so abundant throughout the Arctic regions and the elevated mountains of the north temperate zone, should not have been seen in the southern by any collectors save Dr. Gillies, who is stated (Bot. Miscell. vol. iii. p. 138), to have found a solitary specimen on the Andes of Mendoza, and Commerson. It certainly does not appear amongst the plants of the Colombian Andes that have been published by Humboldt, or more recently discovered by Col. Hall and Professor Jameson. Variable as the species of this genus are in the Northern Hemisphere, they are equally so in the Southern; still, as they seem to present tangible characters, I have availed myself of them here in drawing up the subjoined list of the South American species known to me DEABA, L.

§ I. Racemo inter folia summa sessili.

1. D. cryptantlia, n. sp.; csespitosa, rarnosa, tota incano-tomentosa, foliis imbricatis obovatis obtusis, racemo foliaceo abbreviato inter folia suprema sessili paucifloro, floribus parvis brevissime pedicellatis, petalis spathulatis albis unguieulatis, siliculis longe pedicellatis ovatis incanis stylo brevi terminatis.

Hab. Peru; Cerro Pasco; Mathews (v. 942).

Caules robusti f— 2 imc. longi. Folia uncialia, ramis appressa. Flores ineonspicui. Sepala -i lin. longa, oblonga, obtusa. Petala sepalis vis longiora, apice emarginata. Silicidce 2-i liu. longa? ; seminibus parvis rufo-brunneis.

§ II. Subacatdes v. caulibus caspitosis, pedunculo nudo,fioribns corymbosis, stylo brevi.

2. D. aretioides, H. B. K. ; Nov. Gen. et Sp. Arner. vol. v. p. 77. t, 435.

Hab. Andes of Colombia; Antisana, 10,800 feet; Humboldt and Bonpland. Pichincha, 15,000-16,000 feet; Professor Jameson ; Hartweg, in Herb. Hooker (n. SS5.)

3. D. obovata, Benth. ; Plant. Hartweg. p. 159. n. 885.

Hab. Andes of Colombia; Antisana, 15,000 feet; Hartweg, (n. 885) in Herb.Bentliam.

The species in the Hookerian Herbarium, received as n. 885, from M. Hartweg, is the same as Professor Jameson's B. aretioides and Humboldt's figure ; but it is not the B. obovata, n. 885, of Mr. Bentham's herbarium, apparently a very different species.

4. D. depressa, n. sp. ; csespitosa, incano-tomentosa, caulibus brevibus prostratis ramosis foliosis, foliis confertis obovato-spathulatis apice rotimdatis integerrimis, racemis abbreviatis folia summa paulo superautibus, sepalis late oblongis dorso pubescentibus, petalis flavis, pedicellis fructiferis elongatis, siliculis latissime ovato-rotundatis acutis, stylo brevi, valvis planiusculis incanis, seminibus majusculis late obovatis.

Hab. Colombia; Chimborazo, 17,000 feet ; Col. Hall.

A D.cryptant/ta, cui proxima, difi'ert statura, racemo elongato, floribusque ter majoribus ; ad D.aretioidem statura accedit, sed folia incana obtusa patentia et laxius inibricata, siliculse minores latiores caulesque prostrati. . With regard to the sections proposed for these twenty species, they are tolerably