150 The Philippine Journal of Science m-i 
leaves) . Flowers apparently white, 5-merous. Calyx cylindric, 
cup-shaped, glabrous, 2 mm long, about 1.5 mm in diameter, with 
5, short, blunt, teeth. Corolla-tube 3 to 4 mm long, cylindric, 
externally glabrous, internally villous, the lobes 5, glabrous, 
oblong or narrowly oblong, obtuse, about as long as the tube. 
Anthers lanceolate, 3 mm long, exserted. Style and stigma 10 
to 11 mm long. Fruit globose, black when dry, shining, peri- 
carp somewhat wrinkled, 2-celled, about 8 seeds in each cell, the 
seeds irregular, more or less angular and compressed, about 2 
mm long. 
Guam Experiment Station 26, November, 1911, at Tolijuice. 
A species manifestly allied to Tarenna asiatica O. Ktze., and appar- 
ently also to Stylocoryne sambucina A. Gray, but distinguished from both 
in being entirely glabrous except for the corolla-tube which is villous 
inside. 
I have followed De Dalla Torre and Harms in adopting Gaertner’s 
generic name Tarenna, for those species which have been described chiefly 
as Webera and Stylocoryne, but I am by no means certain that these 
authors are correct in their distribution of the synonyms between Tarenna 
and Randia. I suspect that Cupi Adanson is the oldest, and therefore by 
rules of priority, the correct generic name at least for those species of 
Webera, Stylocoryne, and Tarenna that have several ovules in each cell. 
It is based entirely on Cupi Rheede Hort. Malabar. 2: 37, pi. 23 which is 
the type of Rondeletia asiatica Linn. De Dalla Torre and Harms, how- 
ever, refer Cupi of Adanson to Randia, but I do not think that Rheede’s 
plate and description justifies this disposition of it, and consider that it is 
rather Tarenna than Randia. Webera Schreber cannot be considered, for 
it is invalidated by Webera of Hedwig, a valid genus of mosses. The 
genus Cupia DC. (1830) is typified by the first species cited, C. corymbosa 
DC., which is based on Cupi of Rheede, and the species is hence a synonym 
of Tarenna asiatica (Linn.) 0. Ktze. King has taken exception to the 
generic name Tarenna on the basis that Gaertner described the flowers as 
4-merous, rather than 5-merous, and he therefore revives the genus Stylo- 
coryne Cav. for those species with 5-merous flowers and several-seeded 
fruits.” However, Stylocoryne Cavanilles is invalid, as it is a manifest 
synonym of the genus Randia. Trimen, however, considers Tarenna 
zeylanica Gaertn. to be an exact synonym of Webera corymbosa Willd.= 
Tarenna asiatica O. Ktze., in spite of the discrepancy in the number of 
floral-parts. 
CUCURBITACEAE 
BENINCASA Savi 
BENINCASA HISPIDA (Thunb.) Cogn. in DC. Monog. Phan. 3 (1881) 
513. 
Cucurbita hispida Thunb. FI. Jap. (1874) 322. 
Benincasa cerifera Savi Bibl. Ital. 9 (1818) 158; Safford 197. 
G. E. S. 157, locally known as condor. 
Cultivated in the tropics of both hemispheres. 
”Journ. As. Soc. Beng. 72 1 (1903) 198. 
