570 THE GRASS FAMILY. 



species under the name of C. riparia (Eng. Bot. t. 579). It is also said to 

 have the minute point on the anthers more distinct : but all these characters 

 appear to be too variable to be relied upon as specific. It grows with the 

 smaller form, and is rather more frequent in Britain. 



LXXXVIII. THE GRASS FAMILY. GEAMINE^. 



Herbs, with stems usually hollow, except at the nodes, and 

 alternate, narrow, parallel-veined, entire leaves, sheathing the 

 stem at their base, but the sheaths are split open on the side 

 opposite to the blade, and usually terminate, within the base 

 of the blade, in a small scarious appendage called a licjuJc. 

 Flowers in spihelets, arranged in terminal spikes, racemes, or 

 panicles. Each spikelet consists usually of 3 or more chafF-like, 

 concave scales or bracts, called glumes, arranged alternately on 

 opposite sides of the spikelet, their concave faces towards the 

 axis ; the 2 lowest glumes usually empty, nearly opposite to 

 each other, and often differently shaped from the others. The 

 succeeding, or flowering glumes, enclose each a rather smaller 

 scale called a palea, usually thinner, and with 2 longitudinal 

 ribs or veins, placed either between the glume and the axis of 

 the spikelet, with its back to the axis, or apparently opposite 

 the glume at the end of the axis. Within the palea, or appa- 

 ' rently between the flowering glume and the palea, is the real 

 flower, consisting usually of 2 minute, almost microscopical 

 scales called lodicules, of 3 (rarely 2) stamens, and of a 1-celled, 

 1-ovuled ovary, crowned by 2 more or less feathery styles. 

 The name of flower, however, is here, as ui other works, gene- 

 rally meant to include the flowering glume and palea. Fruit 

 1-seeded and seed-like, called a grain or cargopsis, consisting 

 of the real seed and pericarp, enclosed in, and often adhering 

 to, the persistent palea, and often also enclosed in the more 

 or less hardened flowering glume. Embryo small, at the base 

 of a mealy albumen. 



Such is the general plan upon which the flowers of Grasses are arranged, 

 but there are many variations which require to be carefully attended to in 

 discriminating the genera of this most natural, but somewhat difficult 

 family. Where the spikelet contains but one flower, its flowering glume 

 and inner palea appear often almost opposite to each other, like an inner 

 pair of glumes within the outer empty ones. Sometimes there are three or 

 even more outer, empty glumes, either passing gradually into the shape of 

 the flowering ones, or one or two, very differently shaped (usually much 

 smaller), are placed between the outer empty pair and the flowering one ; 

 or the axis of the spikelet terminates in one or more rudimentary, empty 

 glumes. Occasionally one flower, either below or above the perfect one, 

 has stamens only, and some exotic species are always moncBcious or 



