200 NOTES ON COMMELINACES, 
which some thirty species have been age on slight differences 
in the shape and hairiness of the leaves, n the pubescence of the 
sepals, and in the length of the panicle. T doubt if there are many 
_— in the genus. ave only been able to examine the 
a few cases, but it seems to me not likely that any good 
spec ats “dist inctions will be got out of the fruit. The primary 
division of this genus om been into (2) with six —— (b) with 
it that 
five stamens, but then it is admitted the mo seats 
6-stamened species (D. eden} has eat only yet mens. 
As to the distinction between terminal and radical infloresoonce 
the typical species D. radicalis sometimes produc ces a term 
panicle; while several of the oe fm icled species ane 
branches boring through the base of one of the lower leaf- sheaths. 
on few, or even a single herbarium example: not knowing this 
genus in the field, - i as regards the species been able to do 
little more ~— my predecessors, reducing a few where the 
materi plentifal, or sera the species appear to have been 
founded os "identically ¢ the same thing. In species-making I have 
endeavoured to hit a an I have made somewhat more than 
entham, a good many less than Hasskarl, makes out of the same 
material. Of the 807 a came pen Ty seventy-one are described 
for the first time; these are y from the Tropical African 
collections of Mann, Schweitifaith, ee Welwitsch. Many of the 
othe ave new names, or — in novel dress ; i learnt 
from Mr. Baker that Dracena triandra, Afzel., is a Palisota; and 
from M. DeCandolle that Polygala sinibaeds Poir. ., is a Floscopa. 
6. Besides the new nee described, there are several remark- 
able new species referred to ; 
melina huillensis, Welw. MS., temless species, with 
numerous large azure flowers ; and Welwitsch collected two other 
striking species allied to this in Angola also 
Aneilema sepalosum, sent from Ukamba in Africa by Hildebrandt, 
is a scapose a allied perhaps to Sicdanaio of Royle, but has 
e sepals 3 in. 
neilema janes described originally by Hasskarl as Diche- 
spermum giganteum, supposed a Smilacina by pee — to bea 
Euaneilema » bu wit a capsule more than 3 in 
oR ERO minuta grows in tufts, the stems 1-2 in. high ; 
Se nin by Uhde in Mexico; a single sheet in the Berlin 
7. For the preparation of the monograph of this order, I was 
able to e <r siere in det; ail the Kew collection, including Wight’s 
private Herbarium, containing the ipsissima exempla from which 
his Icones were donated. At the Linnean Society I examined 
