TRIANDRIA— DIGYNIA. Poa. 159 



erect, variegated with glaucoiis-grcen, purple, and silvery white. 

 Calyx-glumes ovate-lanceohite," very acute, strongly 'keeled,' 

 smooth, except the upper part of the keel • membranous at tlie 

 edges 3 the inner one largest, with a short rib near the keel at 

 each side j the outer single-ribbed. Florets 2 or 3 in the first 

 variety J 4,5, or more, in the second, /3, always longer than the 

 calyx. Outer valve of the corolla ovate-lanceolate, either smooth 

 or minutely downy, hairy at the base; furnished with 5 ribs in- 

 cluding the keel, which, like the 2 nearly marginal ribs, is 

 clothed half way up with close silky hairs, the intermediate ribs 

 being smooth, and often so little prominent as to be discoverable 

 only by holding the glume against the light ; inner valve oblong, 

 slightly cloven, with inflexed edges, often rough at the fold. 

 IScct., according to Mr. Sowerby, of 2 notched scales, but I 

 suspect it to vary in this respect, as it does in size. Styles 

 scarcely any. Stigmas large, feathery, and distinctly compound 

 evep in a dried specimen. 

 (5 has broader leaves, and 4 or 5 Jlorets, even in a wild specimen 

 from Mr. Turner ; in cultivated ones there are often six, in 

 which case their common stalk is hairy, particularly close to 

 each floret ; but this is essentially different from the folded web 

 connecting the florets in other species. The calyx is often 

 broader, and quite ovate, in this variety, but there are imper- 

 ceptible gradations. The two varieties however remain constant, 

 through a long course of cultivation, and I have specimens of the 

 original glauca, raised from seed in Mr. Griffith's garden, quite 

 unaltered. Dr. Wahlenberg observed the leaves of glauca to be 

 often involute wiien dried ; ours seldom exhibit this character. 

 All things considered, I agree with the very able botanist last 

 named in reducing these two grasses, hov/ever different in 

 aspect, to one species ; and am happy to concur also with my 

 valued friend Professor Hooker in the same opinion ; but not 

 in referring P. glauca to the very distinct and well-marked P. ne- 

 moralisy though as Dr. Wahlenberg says it is an intermediate 

 species, he cannot mean a doubtful one, betv.-een the latter and 

 P. trivia lis. 



9. P. nemorcdis. Wood Meadow-grass. 



Panicle spreading, capillary. Calyx-glumes lanceQlhve, ta- 

 per-pointed, each three-ribbed. Spikelets lanceolate. 

 Florets about three, five-ribbed, acute ; silky at the keel 

 and lateral ribs ; hairy at the base, without a web. Sti- 

 pulas very short, notched. 



P. nemoralis. Linn. Sp. PI. 102. Willd. v. 1. 399. FL Br. lOG. 

 E}igl. Bot. V. 18. t. 12G5. Knapp t. 58. Hook. Scot. :iD. Schrad. 

 Germ. v. 1. 301. Host Gram. v. 2. 51. t.7\. Leers 30. /. 5. /'. 3. 

 FL Dan. t. 749. Ehrh. Calam. 5. 



P. angustifolia /3. Huds. 41. 



VOL. f. K - 



