XIV, 3 Beccari: The Pal'ms of the Philippine Islands 301 
speaking the same species is rarely represented by individuals 
offering an absolute uniformity of characters, when inhabit- 
ing different islands, even when these are not separated by a 
wide extent of sea. There are some large species specially 
worthy of remark, for they are so much like A. Catechu, as to 
render it uncertain whether they are to be regarded, at times, as 
simply varieties of that palm or as distinct species; they are 
A. m,acrocarpa, A. Whitfordii, and A. Catechu var. longicarpa. 
The first two are really forest plants, and apparently also the last 
is a forest plant. Areca Catechu var. silvatica possibly repre- 
sents the true wild plant of that widely cultivated species (of 
which the true place of origin has not been ascertained), for in 
the Philippines various forms of Areca occur so closely related 
to A. Catechu as to afford good ground for belief that it was in 
those Islands that A. Catechu finally assumed the specific char- 
acters it now exhibits. In support of this opinion I would ob- 
serve that in no other part of southern and eastern Asia or 
Malaya is any species of Areca to be found which in any way 
approaches A. Catechu in specific characters, whereas in the 
Philippines an entire group of species exists closely related to 
it. Areca parens is another fine species, endemic in the Phil- 
ippines, which, although endowed with characteristic and easily 
appreciable diagnostic characters, evidently belongs to the group 
of A. Catechu. 
The fruits of Areca I pot and A. camarinensis in their prin- 
cipal characters and general appearance are also very much like 
those of A. Catechu, but these two species belong to the group 
that produces female flowers, and then fruits, densely grouped 
around the main axis of the spadix, and are therefore related 
to the species for which was established the subgenus Balano- 
carpus, which includes A. glandiformis, of the Moluccas ; A. ma- 
crocalyx, of the Papuan Islands, and a few others. 
Areca costulata is no doubt related to A. (Mischophloeus) 
panicidata Scheff., a palm growing in northern Celebes and in 
the Moluccas; not only on account of its male flowers having 
the cyathiform calyx slightly 3-toothed and with a solid base, 
but also in its general habit. Areca Caliso is somewhat related 
to A. costulata, from which, however, it differs by the calyx of 
the male flower being divided into three small distinct sepals; 
its nearest allied species is probably A. oxycarpa, of Celebes. 
Areca Vidaliana and A. Hutchinsoniana are of the type of the 
very widespread A. triandra, not only on account of the male 
flowers being inserted in pairs on unilateral notches of the flor- 
iferous branches, but also because the male flowers have six 
164391 4 
