122 - The Philippine Journal of Science 1915 
others the forms are separated only by slight, although ap- 
parently constant characters. In habit our species are under- 
shrubs, shrubs, vines, and trees, and they may be entirely gla- - 
brous, or from slightly to densely pubescent. They occur to 
a limited degree in the open settled areas, but chiefly in the 
forested regions, and extend from sea level to an altitude of 
nearly or quite 2,500 meters. A very few species are of wide 
distribution in the Archipelago, but the bulk of them are of 
decidedly local occurrence, and most of them are endemic. The | 
genus is so large that it is decidedly inconvenient properly 
to consider the numerous forms, and after a careful consider- 
ation I have deemed it expedient to recognize the genus 
Grumilea Gaertner for those species which have seeds with 
ruminate albumen. Following Valeton and Ridley, I consider 
it also expedient to segregate the genus Streblosa Korthals, the 
species falling in this genus being entirely different from those 
of Psychotria and Grumilea in habit, in their minute fruits, and 
in their constantly lateral inflorescences. In consideration of 
the recognition of Grumilea as a distinct genus, it has become 
necessary to rearrange our numerous Philippine forms that have 
previously been referred to Psychotria. In the following ar- 
rangement I have limited Psychotria to the species having seeds 
with the albumen not at all ruminate, and using this character 
as a distinguishing one we still have plants of very diverse 
habit and appearance, undershrubs, shrubs, vines, and small 
trees, with fruits, pyrenes, and seeds varying from entirely 
smooth to deeply sulcate or ridged. The following list includes 
the species that, from the character of the albumen, can be 
referred to Psychotria proper. 
In the course of this study I have been able to eliminate from 
the Philippine flora the genus Coelospermum, C. ahernianum 
Elm. being Psychotria leptothyrsa Miq., as previously deter- 
mined by Valeton by citation of a cotype of Elmer’s species, and 
two species of Amaracarpus, A. pubescens Elm., non Blume 
being Grumilea amaracarpoides Merr., and A. longifolius Elm. 
being a synonym of Psychotria linearis Bartl., as already noted 
by Mr. Elmer. 
PSYCHOTRIA LIANOIDES Elm. Leafl. Philip. Bot. 3 (1911) 1029. 
MINDANAO, Elmer 11645. 
The alliance of this species is with P. diffusa Merr. 
PSYCHOTRIA OVALIS Elm. Leafl. Philip. Bot. 3 (1911) 1023. 
MINDANAO, Elmer 10567. 
