234 
PACIFIC SCIENCE, VoL XIX, April 1965 
anthers; column 6-9 mm long, anther bearing 
on upper 2 A’ s; free filament tips 1-2 mm long; 
anthers 1.6-2. 2 mm long, narrowly lanceoloid, 
auriculate at base, bearing at the apex a subu- 
late projection of the connective 0.6-0.9 mm 
long. 
HOLOTYPE: Singapore, Kampong Tanjong, 
Labrador, cult, for making of mats, 11 March 
1937, E. J. H. Corner 32,763 (sing). Iso- 
type (K)! 
SPECIMENS EXAMINED: all in staminate 
flower. 
Java: Hort. Bogor (CAL, L); ex hort. Bogor 
(CAL); cult. Mus. Heyne, Batavia (l); cult. 
Museum tuin no. 24 (l); Java, W. H. de 
Vriese ( L ) ; without data ( CAL ) ; without data, 
"puedak” (CAL). 
India: H. B. C. (Hortus Botanicus Calcutten- 
sis), (CAL, K); H. B. C, Griffith 6,363 (k); 
ditto, Wallich 8,389 (k) . Without Locality: 
1867, Teysmann (l) . 
DISCUSSION: The above specimens from Java 
apparently represent P. moschatus Rumph. ex 
Miq., and those from India were the basis of 
P. inermis Roxb. 
It is apparent from the habit and from the 
staminate flowers that this is a member of the 
section Pandanus. The main problem is to de- 
cide on the species. It is a plant cultivated by 
the native peoples from southeastern Asia to 
the extremes of Polynesia. It is the preferred 
sort for the making of mats, due to its bland, 
unarmed leaves. It is mostly sterile, perhaps due 
to the weakening effect of continual leaf har- 
vests. The natives say that it never bears fruit. 
Rarely, larger, older plants bear male flowers, 
as does the Corner 32,763 specimen here de- 
scribed. So, it is a variant, always propagated 
vegetatively. It appears to have started as a bud 
sport on a staminate plant of some species of 
the section Pandanus . The staminate flowers of 
all species in this section are very similar, but 
only a small percentage of them are described 
and completely known. The leaves, lacking the 
spines, are not sufficiently distinctive to help 
in the specific determination. Earlier botanists 
described it as a species, then as a variety, 
under several names, and attached it to several 
species. Upon the assumption that all littoral \ 
Pandanus belonged to a single species, it was j 
attached to either P. tectorius Soland. or to P. 
odoratissimus L. f. 
It is evident that it is not a species. Prob- 
ably it is not a hybrid. With little doubt it Is 
a bud sport obtained millenia ago from some 
wild species. Its current classification should 
surely be that of a cultivar, and as such it 
should be attached to a species. 
The author now attaches this cultivar to P. 
spurius Miq., a species also native to Amboina. 
It is of the same section Pandanus, and has 
been well figured and described. Though long 
misunderstood by botanists who had no speci- 
mens, it is now represented by several collec- 
tions from Amboina in the herbarium at ; 
Firenze. In a subsequent part of this revision j 
the writer will present new figures and an ex- ] 
panded description, in order to make P. spurius | 
better known. We have not proved that the cv j 
'PUTAT’ was derived from P. spurius, but they • 
are of the same section, occur in the same I 
region, and on the same shores and inhabited 
lowlands. The Ambonese people have long used j 
the leaves of P. spurius for matting. If, long j| 
ago, they found a spineless bud sport, it is 
likely that they would have preserved it and ! 
multiplied it by stem cuttings. 
The name here given to this cultivar is j; 
PUTAT,’ the vernacular name on Java ( fide j 
Rumphius), now as "pudak,” the Sundanese ! 
vernacular name ( fide Koorders). 
The classification of this cultivated variety is ;■ 
difficult, and previous botanists have given It jl 
many "placements. The first good account and i 
naming of this plant was by Rumphius ( 1743) jj. 
who gave it the alternative names F. moschatus j| 
seu laevis, but he himself actually accepted F. 5 
moschatus. In post-Linnaean literature this was | 
validated as F. moschatus Rumph. ex Miq. 1 
( 1855 ) ; and as F. tectorius Soland. var. moscha - j| 
tus ( Rumph. ex Miq. ) Merr. ( 1917 ) . 
The alternative name given by Rumphius jj 
was also republished and validated as F. laevis 
Lour. (1790); as F. laevis Rumph. ex Willd I 
( 1805 ) ; as F. laevis Rumph. ex Kunth ( 1841 ) ; Ij 
as P. laevis Rumph. ex Solms (1878); as P. : i 
tectorius Soland. var. laevis (Kunth) Warb. 
(1900); as F. tectorius Soland. forma laevis ji 
i 1 
