10 TRIURIDEZ AND BURMANNIACEJE 
uniseriate, even when the number of its divisions is double that of the 
stamens, as well as in the apparently simple straight embryo. 
The genera proposed by Miers are five, a number which subsequent 
discoveries would tend rather to reduce than to enlarge. The two first, 
Triuris and Hexuris, each with a single species, have dicecious flowers, 
a terminal style, and the lobes of the perianth equal in number to that 
of the stamens, three in Zriuris, six in Hewuris, The three others, 
Soridium, Sciaphila, and Hyalisma, closely resemble each other in habit, 
and are well distinguished from the two former by their moncecious 
flowers (the upper ones male, the lower female), their lateral, almost 
basal styles, and the lobes of the perianth twice the number of the 
stamens. In Sciaphila, the principal genus, now consisting of four 
Asiatic* and four American species, the perianth has six divisions, with 
generally, if not always, three stamens.t The Brazilian Soridium, and 
the Cingalese Hyalisma, each consisting of a single species, differ from 
Sciaphila solely in these numbers being reduced to four and two in 
Soridium, and increased to eight and four in Zyalisma. But as one of 
the new species of Sciaphila, described below, bears occasionally tetra- 
. merous male flowers, precisely like those of Soridium, it may be a doubt 
= whether the three genera ought not to be re-united into one, under 
Blume’s original name of Sciaphila. — 
Mr. Spruce gathered a variety of Triuris hyalina, Miers, with smaller 
.* These Asiatic species are :— 
= Ig nana, Bl.; racemo paucifloro, perianthii segmentis lanceolatis imberbibus, 
: stylo ovario pluries longiore.—From Java, and apparently the same species, gathered 
in Khasia by Drs. Hooker and Thomson. 
2. 8. tenella, Bl.; racemo multifloro, perianthii segmentis lanceolatis apice intus 
3, stylo clavato-penicillato ovarium vix superante.—From Java and from the 
if, as there is every reason to believe, the S. maculata, Miers, and S. 
pa BL, are co 
3. S. erubescens, Miers; racemo multifloro, perianthii segmentis lato-lanceolatis 
 imberbibus, stylo clavato-penicillato ovarium vix superante.—From Ceylon. 
— 4. S. secu a, Thwaites, MS.; racemo paucifloro, pedicellis secundis, perian- 
thii segmentis longe subulato-acuminatis imberbibus, stylo clavato-penicillato ovarium. 
vix t ridet Ceylon (Thwaites). The flowers are considerably larger than 
~ + Both Miers par^ Blume ascribe indeed to Sciaphila six stamens, but in all the 
ens I have examined, both Asiatic and American, I find but three ; and that is 
= eet number represented in Dr. Hooker's beautiful dissections of the Khasya 
species, and mentioned by Dr. Thwaites in his manuscript descriptions taken from 
fresh Cingalese specimens. It appears that, in some instances at least, the two lobes 
of the anthers, which, before they open, appear to be disjoined, have been mistaken 
for two distinct anthers, In the old flowers the two cells become confluent at the 
