2 MR, C. B. CLARKE ON 
including all the new ones (11), belong to one subgenus, Cari- 
candra. This subgenus includes, in the World, 125 species ; of 
these, 19 recede from the usual habit of the subgenus by the 
stem having only 1-5 spikes; these 19 show affinities to species 
in various other subgenera of Carex. The main, typical, Cari- 
candre, 106 in number, with numerous spikes, sometimes called 
the Section “ Indice,” form a natural group of species, very 
homogeneous, and therefore very difficult to separate into species. 
They are an essentially Tropical group, of which 77 occur in 
South-east Asia, 16 in Tropical Africa, 13 in Tropical America ; 
there is no Carex in Palearctica or Nearctica that can be com- 
pared with them ; the culm bears numerous (10-200) spikes. 
In the Flora of British India, the group Indice contains 
52 species out of a total of 142 Carew in India. That the Malay 
Flora should possess 36 species of Indice, out of a total of 54 
Carex, is therefore what might have been expected. The group 
Indies, however, becomes much scarcer in China, and only a few 
reach Japan, where other sections of Carex abound. 
The Central American and Tropical African species of the 
group Indies are so closely allied structurally to the South-east 
‘Asiatic, that I have not found it possible to give absolute diag- 
noses to distinguish them ; thus Boott made a Madagascar plant 
conspecific with Carew bengalensis, Roxb. At the same time, I 
do not find that the closely-allied Carea of this group from 
different continents match; I have therefore treated these species 
geographically. Several of the species described below as new 
were catalogued in Dr. Stapf's Kinabalu plants (Trans. Linn. 
Soc. ser. 2, Bot. vol. iv. [1893] p. 246) by me under old names : the 
corrections in the present paper are attempts to define geo graphic 
subspecies as species; that is to say, the Kinabalu Carex (group 
Indies) do not exactly match the closely-allied Indian. 
The group Hemiscapose, as to the plants typieally belonging 
to it, appears very plainly defined; there are at the very base 
and on the sterile tufts long, well-developed leaves; the stems 
have the peduncle-holding bracts short, and the stem-leaves, if 
any, are similar to and hardly longer than the bracts. Many 
species are thus definitely hemi-seapose. But the character in 
many other species is less marked, and we have finally a large 
number of species which may be put in some other group alto- 
gether or may be attached to the Hemiscapose. Dr. Theodore 
Holm believes that he is always able to locate the species in its 
