bo 
Or 
g P de P porn 
eine 6 vel 5, 4,3: quorum 8 aut 2 perfecta similia, cetera sterilia difformia. Peduneuli ad dichotomas bracte- 
ati. Capsula regulariter trilocularis elliptica apice acuta. Semina in quoque loculo unica serie superimposita, vel solitaria. 
Sect 1. Ovula 9 vel plura. Semina (nisi casu) 9 vel plura. 
2. Ovula 6. Semina (nisi casu) 6. 
3. Ovula 8. Semina (nisi casu) 3. 
This genus is generally easily distinguished from Commelyna by the inflorescence being subpaniculate, and the peduncles 
entirely exserted from the bracts at the branching of the panicle. But in Anilema spectabile (which has two perfi ct stamens and monoli- 
form hairs on the filaments) the bracts are sub-cucullate and the peduncles included therein. This is the only Anilema, known to me, not 
distinguishable at once from Commelyna by its inflorescence: but in this species the capsule and seeds equally as the stamens leave no 
doubt that it belongs to Anilema rather than to Commelyna. 
In the species of Anilema with 2 perfect stamens, there are often only 3, sometimes only 2, sterile difformed stamens : but 
in large specimens of the same species sometimes 4 sterile stamens are seen. So in the species with 3 perfect stamens, the barren 
stamens are 2 or 3 often on the same panicle. The pains which have been taken in describing these sterile stamens seem to me ill- 
bestowed. 
The number of fertile stamens in each species is in general tolerably constant, But Anilema lineolatum, which has gene« 
rally 3 perfect stamens, has one of these sometimes smaller, sometimes altogether absent. So in Animela protensum the fertile stamens 
are normally 3 but not infrequently 2 only. 
In specific distinctions, stress has been laid on “ filamenta barbata,” as opposed to * filamenta imberbia.” These monoli- 
form hairs often are very dflicult to find in the bud, being greatly developed at the time of expansion of the flower. Thus R. Brown as- 
signs to Anilema vaginatum bearded filaments (as L have always seen them) whereas Wight found them beardless, though it is clear from 
his figure that he had got the true plant. 
On the whole I find that, as in Commelyna, the most constant characters are furnished by the capsule and seeds. 
Dichwspermum of Wight differs by having the seeds in two rows in each cell. Hasskarl in the last revised list of genera in 
his Commelynacez says “ gemmule orthotrope 9.seriatw.? In Wight Icones t. 2076, the capsule is shewn with seeds in one row, the 
ovary with ovules in two rows. I possess full specimens of the so-called Dichwspermum ochraceum : they have the seeds in one row. 
The fact is that both in Anilema and in Dicheespermum the ovules are in two rows : in Anilema the seeds, as in the pod of 
a pea but placed much closer together, appear in one row : in Dichespermum the seeds slightly dovetail but form two distinct rows. 
Tab. XIV. fig. 6. is an ideal horizontal section of the ovary of an Anilema. The 6 terminal points marked H are the 
points of attachment of the ovules so that (as is generally the case) there are two confluent placental lines which in Anilema are at the 
inner angle of each cell, The ovules are thus in two rows, the points of attachment along each placental line being alternate, as is intend- 
ed to be shewn in Tab. XIV. fig. 7, which represents the appearance looking towards the axis of the ovary when the back of one cell has 
been removed. As shewn correctly in Wight’s pictures, when the ripe valve dehisces from the capsule of an Anilema, the seeds remain 
on one side entirely : this is because the seeds in ripening grow so as to fill the cell closely, and each seed fits accurately to the next. The 
capsule valve ultimately dehisces loculicidally at the points E, F in Tab. XIV. fig. 6, and the whole of the septum to the very axis comes 
away with it. Hence in removing one valve of an Anilema capsule, two rows of seeds are generally seen as in Tab. XIV. fig. 9: but the 
seeds are nevertheless here strictly in one row in each loculus and the capsule depicted in Tab. XIV. fig. 9, contains 15 seeds only. 
I suspect mistakes have occurred : notably in the case of the so-called Dichespermum ochraceum which is a true Anilema, 
The capsule in several species of Anilema becomes more or less lustrous or lead-blue : Aclisia however is well separated by 
the seeds being in two or more rows in each cell of the capsule. j 
Dictyospermum of Wight is a very convenient subdivision of the genus if the character is brat to include all the — 
ed species : é. ¢. if it is made equivalent to my sect. 3; but Wight has by his name intended to narrow Dictyospermum to those species 
which have lacunar rugose seeds. 
Species infra deseripte. 
Sect L 
J. Anilema scapiflorum. 
2. A. lineolatum. 
8, 2 divergens. 
4. A. Hookeri. 
By dtl nanum. 
6, A. triquetrum. 
7 A. spectabile. 
Sect IL. 
8. A. nudiflorum. 
9, A. ensifolium. 
Sect IH. 
10. A. yaginatum. 
11. A. protensum. 
12, A. ovatum. 
