NEW OR NOTEWORTHY PHILIPPINE PLANTS, VII. 249 



Panicum nodosum Kuntli. which is closely allied, differs mainly in being much 

 smaller, Avith shorter, narrower leaves, and much smaller panicles. The type of 

 Kunth's species was from the Philippines, P. nodosum being based on P. multinode 

 Presl, non Lam. Balansa, apparently assuming the Indian form to be typical 

 Panicum nodosum Kunth, has distinguished the smaller form, apparently typical 

 P. nodosum Kunth, as P. ouonhiense. 



I am disposed to refer to Panicum malaharicum (Linn.) Merr., two specimens 

 from the Philippines, Palmas Island, Merrill 5367, and For. Bur. oSJfO Merritt, 

 from Mindoro. The former was previously referred by me to P. nodosum Kunth. 



In a broad sense Panicum malaharicum might include P. nodosum Kunth, but 

 I am disposed to consider the small form described by Kunth as wortlij' of specific 

 rank. 



Panicum cordatum Biise in Miq. PI. Jungh. (1854) 376. 



P. luxurians Merr. in Philip. Journ. Sci. 1 (1906) Suppl. .359, cum syn., non 

 Willd. 



Luzon, Province of Union, Bauang, Elmer 5595: Province of Bataan, Lamao 

 River, Williams 134, Elmer 6135: Province of Bulacan, Norzagaray, Yoder 1, 163. 

 CuLioN, Merrill 55 Jf. 



This species was previously determined by me as Panicum luxurians Willd., 

 but is apparently not that species. I have examined the material in the Will- 

 denow Herbarium so named, there being two specimens, one marked "Klein Ind. 

 1799" which is apparently a form of Panicwrh prostratum Lam.=P. reptans Linn., 

 the other from St. Vincent, collected by Bory, which may be the same as P. 

 caespitosum Sw. Willdenow, Enum. 1 (1833) 109, gives as the only locality for 

 the species '"Luzonia," an apparent error on his part. 



The specimens cited above closely resemble Panicum montanum Roxb., but can 

 l)e readily distinguished from that species by the first glume being as long as the 

 spikelet. The Philippine .material differs from Javan cordatum in its relatively 

 narrower and less prominently cordate leaves, but does not seem to be specifically 

 distinct from that species. 



Panicum reptans Linn. Syst. Nat. ed. 10, 2 (1759) 871; Hitchc. in Contr. 

 U. S. Nat. Herb. 12 (1908) 119. 



Panicum prostratum Lam., 111. 1 (1791) 171; Kunth Enum. 1 (1833) 89; 

 Steud. Syn. 1 (1854) 61; Hook. f. Fl. Brit. Ind. 7 (1897) 33; Merr. in Philip. 

 Journ. Sci. 1 (1906) Suppl. 355. 



The oldest name for this common and widely distributed species is Panicum 

 reptans Linn., as shown by A. S. Hitchcock 1. c, who has examined the type in 

 the Linnean herbarium. 



Panicum paludosum Roxb. Hort. Beng. (1814) 6, nomen ; Fl. Ind. 1 (1820) 

 307. 



Panicum proliferum Hook. f. Fl. Brit. Ind. 7 (1897) 50; Merr. in Philip. 

 Journ. Sci. 1 (1906) Suppl. 359, non Lam. 



Mr. A. S. Hitchcock has examined the type of Lamarck's Panicum proliferum 

 in the Paris Herbarium, and finds it to be identical with Panicum miliare Lam.,' 

 and has accepted P. dichotomiflorum Michx. as the earliest valid name for the 

 species, commonly identified bj' American authors as P. prostratum Lam. I am 

 now of the opinion that the species of tropical Asia and >\Ialaya is distinct from 

 the American form, and have here adopted Roxburgh's sjiecific name for the 

 oriental form, as it seems to be the earliest valid one. 



« Contr. U. S. Nat. Herb. 12 (1908) 147. 

 85754 4 



