Page 455: Revision of Pandams, 28 — St. John 
419 
evident to the base of the free part of the 
phalange; carpels ”9-11,” but 10 on the single 
lectotype, their apices ovoid, the marginal ones 
with a small, oblique, deep, concave platform; 
central apical sinuses 4-5 mm deep, straight, V- 
shaped; stigmas 1-2 mm wide, elliptic to reni- 
form, centripetal, very oblique, brown, papil- 
lose; proximal sinus a short slit, running y Q — 
^4 way to valley bottom; endocarp apparently 
supramedian, but the single phalange was not 
cut for examination. 
lectotype: Australia, "(T.) v. v.” and his 
manuscript adds: Sandy Cape, Hervey’s Bay, 
[north end of Great Sandy I., Queensland], 
July 31, 1802, R. Brown 5, 7 99 A, the phalange 
7.6 cm long, with ovoid carpel apices (bm). 
discussion: P. pedunculatus is a member of 
the section Pandanus, though Brown’s first 
identification of it was as P. odoratissimus L. f. 
His manuscript, obviously written in the field, 
since he recorded, "lect: July 31: descr. Augt. 
3,” is of 23 lines and is reasonably detailed, 
but his published diagnosis was of 12 words, 
of which 8 were descriptive. Of these the only 
significant ones were "phalangibus 8-18-locu- 
laribus : apice conico-lobato; ...” The species 
was accepted and given a vague, generalized 
description by Bentham & von Mueller (FI. 
Austral. 7:149, 1878). It was likewise accepted 
by I. B. Balfour (1878), Warburg (1900), 
and Martelli (1913). The latter (Webbia 
4(1) : t. 1, fig. 1-7, 1913) extended the range 
from Australia to the Loyalty Islands, the 
Solomon Islands, and the Bismarck Islands, and 
illustrated the fruits from each area. Each has 
a different and distinctive appearance. The 
Australian one was actually drawn, not from 
Brown’s type, though he had studied the British 
Museum collections, but from a specimen col- 
lected by F. von Mueller, labeled merely 
"Australia.” This, as represented in Martelli’s 
herbarium, is a single phalange 5.7 cm long, 
4.9 cm wide, broadly obovoid, the scarcely 
convex apex nearly as wide as the middle; 
carpels 25 and their apices mostly conic, but a 
few outer ones ovoid-conic, the central apical 
sinuses 5-7 mm deep. The present writer does 
not consider it conspecific with P. pedunculatus, 
nor should it be used as a standard to represent 
that species. It is here redetermined as P. Blakei 
St. John. The three collections cited by Martelli 
from Melanesia are also extraneous, but do not 
need a detailed discussion here. 
Martelli’s classification of members of the 
large and difficult section Pandanus fluctuated 
from his early view that it contained only a 
widely dispersed littoral species which could not 
be called P. odoratissimus L. f., as that name 
must be used for an aggregate species; to the 
view that there were many separate species; 
to the view that there was one major species, 
P. tectorius, under which many varieties could 
be recognized; to the view that this should be 
called P. odoratissimus L. f.; and then to reverse 
swings to one of his previous views. Martelli 
never printed any general key to the species or 
taxa he accepted. He long studied Pandanus and 
published on it over a span of 30 years, and his 
work was good, but in it there are certain in- 
consistencies. As regards P. peduncidatus, the 
writer finds it to be exclusively Australian, and 
he has not seen any more recent collections 
to match the type, and no one has taken it 
again from the type locality. 
Bentham wrote (FI. Austral. 7:149, 1878), 
"I found no specimen in Brown’s Herbarium, 
. . .” but he did find one without label which 
he identified as P. odoratissimus L. f. The 
printed label form for Brown’s collections was 
prepared, and collection (or species) numbers 
were assigned, by J. J. Bennett when he acces- 
sioned the herbarium into the British Museum. 
Hence, it is likely that the specimen here 
accepted as the type of P. peduncidatus was the 
one which Bentham had called P. odoratissimus. 
The collection Brown 9 ,7 99 A is of three 
phalanges. First, the one just described and 
here made the lectotype. Second, a phalange 
4.7 cm long, 5 cm wide, 4 cm thick, subglobose, 
broad-based; carpels 10, their apices oblate 
pyramidal and the marginal ones with a small, 
concave distal platform; central apical sinuses 
2-3 mm deep, and the valleys very wide V- 
shaped. Third, is a transverse median section 
1 cm thick of a phalange with 14 carpels which 
seems to be identical with no. 2. These last two 
are surely different from the lectotype, and do 
not agree with the details of Brown’s diagnosis, 
"phalangibus . . . apice conico-lobato; basi sub- 
angustata.” Hence, they are here excluded. They 
may represent Brown’s collection from En- 
