Revision of the Genus Pandanus Stickman 
Part 1. Key to the Sections 
Harold St. John 1 
For THIRTY YEARS the writer has lived and has 
made plant collections on the tropical Pacific 
islands, the home of a great part of the genus 
Pandanus ( Pandanaceae ) . His gatherings of the 
genus are numerous. 
In 1955 and 1957 grants were received from 
the National Science Foundation of the United 
States, to make possible extensive exploration 
for this genus in the Pacific, for the employment 
of an assistant, and for research investigations. 
Most of the species are trees with elongate, saw- 
toothed blades, and large, fleshy fruits that are 
difficult to dry. Many of them have unpleasantly 
spiny trunks, branches, and even roots. Conse- 
quently, the collecting of specimens is difficult 
or hazardous, and the drying of them is a long 
and burdensome process. But few of the botani- 
cal explorers in the Pacific have had the interest 
or the endurance to make extensive collections 
of Pandanus. Most of the others, when collecting 
their first specimen, have been lacerated by the 
saw-like leaves or wounded by the spines of 
the trunk. They have then uttered an oath and 
sworn never to collect another Pandanus. As a 
result the Pandanus population has been well 
sampled in only a few areas. 
Outstanding among previous explorers was 
Prof. Ryozo Kanehira of Kyushu University, 
who revealed much of the wealth of Pandanus in 
the Caroline Is., Marshall Is., and New Guinea. 
1 B. P. Bishop Museum, Honolulu 17, Hawaii, 
U. S. A. Manuscript received June 12, 1959- 
Since 1955 the writer or his assistant, Ben- 
jamin C. Stone, have made lone explorations 
for Pandanus in the Ryukyu Is., Marianas Is., 
Caroline Is., Marshall Is., New Guinea, Bis- 
marck Arch., Solomon Is., the New Hebrides, 
and the Hawaiian Is. In almost every one of 
these islands or archipelagos were discovered 
from two to five times as many kinds of Pan- 
danus plants as were known there from earlier 
explorations. 
The writer envisions that his published stud- 
ies will form a revision of the genus, or at least 
of its species occurring in the Pacific. As small 
portions of it are finished, they will be pub- 
lished as numbered parts, with a consecutive 
paging, under one general title. In this way the 
finished parts can be made available early. 
Many of the present sections of the genus are 
those groups recognized as genera by Gaudi- 
chaud. When only a few species were known, 
many of these groups seemed truly good gen- 
era. With the increase in the number of species 
to nearly 600, the "generic” distinctions have 
diminished or vanished. Hence, the writer con- 
curs with the judgement of nearly all recent 
investigators that Pandanus is one very large 
genus that contains numerous well-marked sec- 
tions. There is nothing to be gained by classing 
them all as subgenera, so the writer follows 
Warburg ( 1900) in dividing the genus into sec- 
tions. More sections have been added recently, 
especially by E. G. R. Pichi-Sermolli (1951) 
in his study of the species of Madagascar. 
224 
