.ALAllA^Tia PHlLimNBNSiA, II. 203 
r 
in 5's, the pedicels are intermediate in length between those of the other two 
groups^ and are nearer the type in thickness^ the inflorescence is usuully lax, the 
bases of the petals rather conform to the second type. 
The plants, in spite of very considerable individual variations, can not be 
distinguished by any vegetative characters. 
The collections which seoni to conform to the first type are Cum\n(j ICtSG, 112G 
(type), without definite locality, but supporied to be from the IbUmd of Samar: 
Luzon, Province of Ilocos Sur, San Quintin, For. Bur, SO-'fG Klcmme (fruiting; 
probably) : Province of Laguna, Los Bafios, Elmer 8294: Province of Tayabas 
(Infanta), Anoling, For. Bur. GG02 Ko'bhc; between Casihan and Guniaca, For, 
Bur. 10262 Curran; Lucban, Elmer 9105, Atimonan, For. Bur. G70S Kohhc: Prov- 
ince of Camarines, Mount Isarog, For. Bur. 10503 Curran; Pasacao, Ahem 18: 
Ticao, For. Bur. 1087 Clark: and presumably the following sterile specimens 
from Tayabas Province; between Paete, Liiguna, and Piapi, For. Bur. 10135 
Curran; Malicboi, Ritchie; between Dagutun and Pitogo, For. Bur. 1021S Curran. 
To the second type belong the following, all from Mindanao: District of 2am- 
boanga. Sax River, Williams 2196 ; Lake Lanao, Camp Keithley, three sheets, 
Mrs. ClemenSj without number; Province of Misamis, Mount !Malindang, For. 
Bur. 4692 M earns d Hutchinson. 
The linking type includes the following; Mindanao, District of Cotabato, 
Balugu, For. Bur. 6559 Hutchinson; Lake Lanao, Camp Keithley, Mrs. Clemens 
616 and 4 unnnmborod sheets; District of Zamboanga, Tctuan, Ahern 376: 
Basilan, Isabela, For. Bur. 3958 Hutchinson, DcVore d Hoover 97; with only 
insular locality, Hallier. 
Two sheets, Whitford 1684, Aroroy, Masbate, and For. Bur. 7403 Danao, 
Occidental Negros, are unfortunately sterile. 
The stamens in this species are always described as free. They also vary in 
this respect. In the extreme development there is a short but perfectly definite 
androgynophore at the summit of which arise the stamens followed by the stami- 
r 
nodes with a very sliort-staJked ovary; but the ovarian stalk may be entirely 
wanting; more commonly the bases of the filaments form a tube 2.5 nun or more 
in length, sometimes they are entirely, free to the base. I Iiave failed to correlate 
these variations with the types noted above; they more probably indicate age. 
This staminal arrangement is very similar to that of Oreu-ia siyJocffrpa Warb., 
an even commoner Philippine plant, ranging from Benguet in northern Luzon 
to Davao in southern Mindanao. The latter is, however, a true Grewia, having 
capitate cymes panieulately arranged, each head surrounded by sepaloid bracts 
and contaijiing 3 or 4 flowers. The sei)als are separate to the base, nuieh 
exceeding the petals, the anthers are not confluent, and mature flowers show 
a short but definite gynophore, which is not evident in very young flowers. 
It is not contended that Diplodiscus should be united with Grewia, as the 
sepals in the former never split to the base, and the anthers are confluent, even 
in buds. 
Pierre^ has reduced Diplodiscus to Brownlowia, but if the separating char- 
acters cited in the Pflanzenfamilien hold good for the latter, it seems prtferahle 
to hold them apart. 
TRIUMFETTA Plum. 
Triumfetta suffruticosa Blume Bijdr. (1825) 113. 
Collected at an elevation of 900 m at Todaya, District of Davao, Mindanao, 
by R. S. Williams, no. 2591, in flower and fruit, April, 1905. 
Hitherto known only from Java. 
* Fh Por. Cochinchine sub. pi. 130, 
