vaginis striatis breviora ; ligula late elongata, obloiiga, apice fimbriata. Panicula 3-5 unc. longa, contracta v. effusa, ramis scaberulis. Locusta 4-5 lin. longa?, vivipara?, pollicares et ultra. Gluma chartacea?, a?quales, acuminata?, superiore 3-nervi, subcarinata, carina scaberula. Flosculi sub 5, basi appresse araneosi, lana albida. Palea inferior flosculorum superiorum sa?pissirne in folium apice uncinatum. ligula et vagina 5-nervi instructum desineus. Antlierce lineares. Ovarium late obovatum, supra basin contractum, basi squamulis acinaciformibus instructum. Styli. breviusculi, ad basin plumosi.
A very handsome grass, which, perhaps, properly belongs to Poa, though the palea? are so decidedly acuminate that I prefer retaining it under Festuca. The two varieties enumerated are not always constant to the characters assigned to them.
In general appearance this species resembles the British Aim caspilosa, which is frequently similarly viviparous on the mountains, and the modifications the palea? consequently undergo both in these and some other grasses, is a subject well worthy of study. When the inflorescence becomes foliaceous, the palea itself, which is distinctly 5 -nerved is represented by the (equally 5 -nerved) vagina of the leaf; the ligula of the latter holds the position of the membranous and often divided apex of such a palea as that of Aim, whilst the lamina answers to a dorsal awn; or rather, in the case of Festuca Fuegiana, to five awns (such as those of Polypogon Chonoticus, p. 374), united by parenchyma. That the arista of the lower palea in grasses is the produced mid-rib of a modified leaf, is perhaps generally admitted, but the exact relation of the apex of the palea to a ligula is not so evident in all aristate florets, as it is in those where the middle nerve is not percurrent but separates from the palea in the form of an awn. One apparent objection to this view may be adduced in the distinctly articulate awn of Coryneplioriis and Stipa, which may further be supposed favourable to M. Easpail's theory, that the mid-rib of the palea is an axis of developement in cohesion with the bracts ; such articulations are, however, exceptional, and their position I am inclined to consider as indicating the point of union of the leaf with the vagina, where an angle is always observable. Viviparous grasses, too, would be expected to produce constantly additional organs from the portion of the transformed palea beyond the ligula, if M. Easpail's view were correct, but, this, on the contrary, is seldom the case. There is a similarity between the palea of a viviparous grass and the upper bract of each spikelet in some Marisci : for in them the dilated lower portion of the bract, or the true continuation of the rachis, somewhat resembles, without however being strictly analogous to, the lower palea of a locusta, and the uppermost flower is borne in a position, similar to the axle of the ligula on the leaf of a grass.
Plate CXLI. Fig. 1, locusta ; Jig. 2, floret; fig. 3, ovary ; Jig. 4, squamida ; Jig. 5 and 6, viviparous portions of a spikelet ; Jig. 7, palea transformed into a leaf: — all magnified. 2. Festuca Arunclo, Hook. fil. F. Alopecurus, LfUrville in Mem. Soc. Linn. Paris, vol. iv. p. 604. Brong. in Dwperrey Toy. Bot. p. 32. Poa (?) Alopecurus, Kuntk, En. Plant, vol. i. p. 256. Arundo Alopecurus, Gaud, in Ann. Sc. Nat. vol. v. p. 100., et in Freyc. Toy. Bot. p. 409.
Var. £. minor, foliis angustioribus culmo brevioribus.
Var. y. pedahs, glumis et paleis latioribus brevioribusque.
Var. 8. culmo graciliore, panicula sub-nutante, flosculis-saepius basi parce lanatis.
Hab. Falkland Islands, all the varieties forming very large tufts ; on the sea-sand abundant. ; D' Urville, J.D.H. Var. y. Strait of Magalhaens ; Port Gregory, Capt. King.
Next to the Tussock, the present is the largest grass in the Falkland Islands, though, like that plant, it is very variable in size. The largest specimens are tlu-ee or even four feet high, the smaller scarcely one. Though a conspicuous object, its varieties are not always easily recognizable ; for the most prominent characters of the typical state, which are the great size of the locusta?, and the narrow palea? and glumes with slender attenuated apices, are quite fallacious. All my large specimens of var. a have either a minute turbinate ovarium or a small
