Page:The Flora of British India Vol 7.djvu/13



Erect decnmbent or creeping herbs, or in Tribe Bamhusea shrubs or trees. Stem terete or compressed, jointed ; iaternodes solid or hollow. Leaves simple, usually long and narrow, entire, parallel-nerved, with a sheathing base distinct from the blade; sheath split to the base (very rarely entire) with often a transverse hyaline erect appendage {ligula) at the union with the blade, facing the latter. Inflorescence terminal, rarely also from the upper sheaths, consisting of spicate racemed capitate or panicled spikelets. Spikelets of three or more alternate distichous bracts {glumes), of which the two lowest are normally empty, and the succeeding, if more than one, are arranged on an axis {rachilla), and are all or some of them flowering; within eacri flowering glume and opposite to it is an erect narrow 2-nerved scale (palea), the margins of which are infolded towards the glume and enclose at the base the true flower. Flowers uni- or bisexual, consisting of 2, rarelv 3 or 6 microscopic scales (lodicules) representing a perianth, and stamens or a pistil, or both. Stamens 3, rarely 1,2, 6, or very rarely many, hypogynous ; filaments capillary ; anthers versatile, fugacious, of two parallel cells, with no apparent connective; pollen globose. Ovary entire, 1-celled ; styles 2, rarely 3, free or united at the base, usually elongate, and exserted from the sides or top of the spike- lets, clothed witb simple or branched stigmatic hairs; ovule erect, ana- t'opous. Fruit a. seed-like utricle (grain) free within the flg. glume and palea, or adherent to either or both ; pericarp very thin, rarely thick or separable from the seed. Seed erect; albumen copious, mealy; embryo minute, at the base of and outside the albumen ; cotyledon scutelliform, bearing on its face an erect conical plumule, and descending conical radicle.

Genera about 300 ; species estimatfd at about 3O0O, but many are doubtful, and more mere varieties; natives of all climates and regions.

In working up the grasses for this Flora, I find the multiplication of species to have passed all bounds, and their nomenclature to be involved in a corresponding degree. This has aitsen from two principal causes, from authors not taking into account the wide area over -^hich the individual species of grasses range,* and from the imperfection of the descriptions of the earlier and many later authors. It is sixty -two years since Kunth published his '* Agrostographia Synoptica (Tubingen, 1833), which is an uncritical sweeping up of all previously known supposed genera and species, with imperfect descriptions and synonyms. It was succeeded (in 1835) by a second volume, in which a few hundred species of the first volume are very fiilly and accurately described, and valuable notes upon others are added. In 1855 Steudel's retards the student of the Order. Of more recent works on Graminece, three only are of great mark, namely, Mnnro's very able Monograph of the BambusecB (Trans. Linn. Soc. vol. xxvi. (1868) ; Bentham's revision of the genera, Gen. Plant, vol. iii.
 * ' Synopsis Graminum " appeared. It in no respect advanced, and in many ways

hitherto unexplored countries, that novelties amongst the grasses are very few indeed, compared with what occurs in other natural families.
 * It is a fact familiar to every one who examines collections of plants from