Page:The Botany of the Antarctic Voyage.djvu/407

Falklands, etc.] to Parry's 1st Voyage, p. 184.). Since the publication of the last mentioned work it has been universally looked as an Antarctic species alone, and its close affinity with the A. pratensis, of the Northern Hemisphere was never alluded to. The ordinary states of the latter plant have a longer and less hairy spike ; but amongst the varieties of it which occur in North Western Asia, and N. Eastern America, there is one wholly undistinguishable from An- tarctic individuals ; and how far these may be constantly distinct appears very doubtful to me. Mi-. Brown, in drawing up the characters of A. alpinus, alludes to his having gathered Scotch specimens with an arista twice as long as the glumes, such is the case with all the Antarctic ones, and in Trinius's figure of A. alpinus ; but is at variance with Smith's specific character, (founded on Mr. Brown's specimens) and with the ordinary state of the Scotch plant. Mr. Watson, however, has gathered the same aristate variety of-/, alpimcs in Scotland, and has cultivated both forms in his garden. His garden specimens of both states are now before me, the long awned one retaining its characters, and the awns of the common form decidedly elongating under cultivation. The comparative length of the lamina and vagina of the uppermost leaf, is also very variable, even in A. alpinus, these being sometimes of equal length, while in the Antarctic plant the lamina is sometimes considerably the shorter ; and, again, I have examined an European specimen of A. pratensis, in which the lamina is even longer than the vagina. The other characters of A. pratensis, used by Mr. Brown, are those of the glumes being acute, and villous only at the sides; this is the case with the British examples that I have studied, but not with the Siberian, which certainly present intermediate forms between this species, and its Fuegian congener. The Antarctic specimens vary exceedingly in size, from four inches, to two and even three feet high; the culms are generally tumid above the upper leaf and contract gradually towards the panicle; or they are slender, cylindical and terete: the lamina of the upper leaf is occasionally far shorter than at other times, equal in length to, or much longer than its vagina. Spikes nearly cylindrical, 2-3 to 1-i inches long, generally rather more than twice as long as broad, but now and then much narrower. Glumes always more or less villous all over.

Admitting the foliage to afford no specific character between A. alpimcs, A. pratensis, and A. Antareticus, and the length of the arista to be very variable in the first of these, there remains no constant character to distinguish these three ; for between A. Antareticus and A. pratensis the only apparent distinctions lie in the villosity of the glumes, and the form of the spike, differences which do not hold in Siberian specimens of the latter. I have added a plate of the common Falkland Island state of this species.

Plate CXXX. Fig. 1, glumes and floret; jig. 2, floret removed from the glumes; fig. 3, pistil : — all magnified-

2. PHLEUM, L.

1. Phleum aljnnum, Linn. Sp. PI. p. 88. Banks et Sol. in Bibl. Banks. Engl. Bot. t.519. P. Haenkeairam, Brest, Eel. Hani. vol. i. p. 245. Nees, in Nov. Act. Acad. vol. xix. Sujopl. p. 140.

Hab. Strait of Magalhaens ; Port Famine and Port Gregory, Copt. King. Good Success Bay, Banks and Solander.

This species, which is associated in the mountains of Scotland with Alopeeurus alpiniis, also accompanies that plant in the southern regions. It has been gathered by Mr. Bridges, on the east side of the Andes of Chili, at an elevation of 6-7,000 feet; and also on the Cordillera of Mexico by Linden, and by Galeotti on the Peak of Orizaba, at an elevation of between 10 and 12,000 feet.

3. MUHLENBERGIA, Schreb.

1. Muhlenbergia rai-iflora, Hook, fil.; rigida, glaberrirna, panicula efl'usa pauci- sub 10-flora, glumis subsequalibus enervibus flosculo paulo brevioribus, palea inferiore lanceolata coriacea basi glaberrima in aristam longissimain rigidam scaberulam desinente superiorem breviorem amplectante, culmo foliato, foliis rigidis setaceis marginibus involutis. (Tab. CXXXI.)

Hab. Cape Tres Montes ; Patch Cove, 2,000 feet, C. Barwin, Esq.