THE SCITAMINEuE OF THE PHILIPPINES. 163 



Staminodes similar but more elliptic. Lip short, oblong, cuneate, widely 

 retuse, the terminal points obtuse. Filament fairly stout. Anther-spurs 

 4, triangular, acuminate, equal. 



Luzon, Province of Cavite, Mendez Nunez, Bur. Sci. 1455 Mangubat ; Mara- 

 gondong, Merrill J/168: Province of Rizal, San Mateo, For. Bur. 1846, 3259 Ahern's 

 collector; Bosoboso, Merrill 2844 •' Province of Pampanga, Mount Arayat, (Warburg 

 14069, fide Schumann sub G. heterobractea) , Merrill 4232: Province of Laguna, 

 Jalajala, (Meyen, fide Schumann sub G. ectobolus) : near Manila, (Barthc, fide 

 Gagnepain sub G. Barthci) : Province of Pangasinan, Merrill 2874. Panay, San 

 Jose, Yoder 14- 



Amboina, Celebes, Batchian, New Guinea and the Bismarck Archipelago; also 

 as a weed of cultivation in Penang, Singapore and Java. 



It is quite impossible to break up this species into the various "species" 

 founded by Schumann and Gagnepain. I can not find a reliable character in any 

 of the Philippine plants that I have seen, which justifies separating specifically 

 any of the forms. G. ectobolus Schum. has for its chief character the production 

 of bulbils in the axils of the stem leaves, a character which I have also seen in 

 typical G. marantina introduced by Javanese from Java into Penang. G. Barthei 

 Gagnepain, which I take to be the typical Philippine form, has broader leaves 

 and a stouter stem than has the form that occurs occasionally as a weed in 

 the Straits Settlements. 



There is a considerable degree of variability in the pubescence of the various 

 parts of the specimens. In some, at least, short processes arise on the corolla- 

 tube and lobes, hardly long enough to be called pubescence. I have long ago 

 pointed out the valueless character of the production or nonproduction of bulbils, 

 for I have hardly met with any Globba in any abundance which does not oc- 

 casionally produce them. G. marantina, which was for many years a weed in the 

 Botanic Gardens in Singapore, and of which Schumann says that he had never 

 seen flowers, eventually produced a few flowers, although almost every plant 

 produced bulbils only for a number of years. Probably G. strobilifera Zoll. 

 (G. Zollingeri Cagnep.) of Java is the same species. 



10. G. ustulata Gagnepain in Bull. Soc. Bot. France IV 1 (1901) 208, pi. 

 7, /. 5-7; K. Schum. in Pflanzenreich 1. c. 158. 



An herb 40 cm tall, with ciliate sheaths. Leaves lanceolate, acumi- 

 nate, petioled, slightly pilose, velvety beneath, ciliolate, 15 cm long, 4 cm 

 wide; ligule 1.5 to 2 mm long, the margins villous. Panicle sessile, 

 dense, 5 cm long. Bracts orbicular or elliptic, 1.2 cm long, black at the 

 tips, bulbiliferous. Flowers two or three together on a short peduncle, 

 black or purple. Ovary smooth, glabrous. Calyx 7 mm long, 3-toothed, 

 the teeth mucronate. Corolla- tube 1.2 to 1.4 cm long, upper lobe with 

 a long mucro 1 mm long. Anther 4-spurred, spurs acuminate. 



Busuanga, Calamianes Islands, (Marche 221 B), in Herb. Mus. Paris. 



Indo-China. 



I should reduce this species also to G. marantina L. from the description, for 

 I have seen no specimens, were it not for the statement that the flowers are 

 black or purple, a most unusual, if not unique character in the genus, and for the 

 mucronate upper petal. 



