m TRIANDRIA— DIGYNIA. Poa. 



Sussex. Mi. Borrer. At Lowestoft, Suffolk, on the low sandy 

 ground between the middle part of the town and the beach, 

 plentifully. 



Perennial. April, May, 



Root a tuft of small, ovate, white scaly bulb*, as truly such as the 

 bulbs of a Lily or Garlick, and, like them, throwing out fibrous 

 radicles from their base ; which happens when the autumnal 

 rains fix them in the moistened sand. Early in spring a dense 

 crop of linear, keeled, slightly glaucous leaves, serrated with fine 

 sharp marginal teeth, is produced, affording a grateful pasturage 

 for cattle, and withering by the time when summer feed abounds. 

 Their sheaths are broad, smooth, partly striated, rather lax, often 

 purplish. Stipulas all lanceolate, acute, of a moderate length, 

 considerably decurrent. Stem 4 or 5 inches high. Panicle ovate- 

 oblong, less spreading and lax than in either of the two last, and 

 scarcely at all zigzag. Spikelets of hardly more than 4 florets, 

 usually of 3, ovate, pale, variegated with green and a violet pur- 

 ple, somewhat shining, externally smooth, except the rough keel 

 of the calyx. Outer valve of the cor. ovate and acute, as in the 

 last, silky near the margin, hairy at the keel with a continuation 

 of the complicated, not very copious, web, which connects the 

 bases of the florets ; inner strongly fringed. Stigmas slender, 

 and, as far as I can judge, simply feathery. 



I find no difference between this grass and what abounds in Ger- 

 many, France and Italy, in the early spring, except that the lat- 

 ter, in the streets of Rome, is usually viviparous, which circum- 

 stance has not been observed in England. Morison's sect. 8. t. 5. 

 /. 14. Ger. Em. 3./. 1, and Barrel. Ic. t. 272. and t. 703./. 1, 

 represent this 5 and possibly /. 2. of the plate last quoted may 

 be the much larger oriental variety, preserved in the Linnsean 

 herbarium, and alluded to in Engl. Bot. as a distinct species. 

 It looks very different, having long and narrow leaves, but is 

 certainly what Linnceus intended by his variety (3. I have other 

 viviparous specimens, which evince a most Proteus-like nature 

 in Poa bulbosa, if they really belong to it ; but this question is 

 here out of place. 



Mr. Knapp'S t. 53 has accidentally the root of some other species 

 subjoined. 



5. P. trivia/is. Roughish Meadow-grass. 



Panicle spreading. Spikelets three-flowered. Florets lan- 

 ceolate, five-ribbed, connected by a web. Stipula ob- 

 long. Stem and leaves roughish. Root fibrous. 



P. trivialis. Linn. Sp. PI. 99. U'illd.v. 1.387. Fl. Br. 103. Engl. 

 Bot.v. 15. t. 1072. Curt. Lond.fasc.2.t.6. Knappt.54. Sincl. 

 2 1 . Hook. Scot. 35. Schrad. Germ. i\ 1 . 296. Host Gram. v. 2. 

 45. /. 62. 



