COMMELYNACEA. 
Dr. Hassxarn’s Monograph entitled “ Commelynacew Indice imprimis Archipelagi Indici” is dated 1870. My paper “On 
“the Commelynacem of Bengal” was read at the Linnean Society of London on 17th February 1870, and was subsequently printed in the 
Linnean Society’s Journal, Botany, Vol. XI. It was written quite independently of Dr. Hasskarl’s far more elaborate work. 
Thave since discovered several new species of Commelynacee in Bengal, and have with larger material worked over the 
Bengal species again by the light of Dr, Hasskarl’s labors. I now publish the result as a corrected and slightly extended edition of my 
paper read before the Linnean Society. The most important correction, viz. the disentanglement of the two genera Pollia and Aclisia, 
Tarrived at by alighting on the two plants called below (in accordance with Dr. Hasskarl’s nomenclature) Aclisia Indica and Pollia 
Aclisia growing together ; but Dr. Hasskarl is entitled to the entire credit of unravelling the confusion which previously existed concern- 
ing these genera. 
Thave had three dried collections to work with, vi -, my own, the Calcutta Herbarium Collection, and the private collec- 
tion of Mr. Kurz, the Curator of the Calcutta Herbarium, which contains many Jaya specimens, some collected in Jaya by himself, 
others issued named from the Buitenzorg Herbarium. 
The books I have been able to consult are: 
Roxburgh’s Flora Indica, Carey’s Edition. 
Roxburgh’s Original Drawings preserved in the Caleutta Herbarium. 
(1) 
(2.) 
(3.) Wight’s Icones Plantarum, Vol. VI. 
(4.) 
5.) 
4.) Kunth’s Enumeratio, Vol. IV. 
5.) Hasskarl in Plante Junghuhniane, published 1852. 
(6.) Miquel’s Flora Batavia, Vol. IIT. 
(7.) Hasskarl on Forrestia, in Regensburg Flora, No. 40, 1864. 
(8.)  Hasskarl, Genera Commelynacearum, in Regensburg Flora, No. 14, 1864. 
(9.) Hasskarl, Commelynaces: Indice, 1870. 
(10.)  Dalzell, in Hooker’s Journal of Botany, Vol. TIT., 1851, 
(11.) Thwaites’ Enumeratio Plantarum Zeylanie. 
(12.)  Bentham’s Flora of Hong-Kong. 
(13.) Edgeworth, in Linnean Society’s Transactions, Vol. XX. 
(14.)  Drury’s Hand-book of the Indian Flora, Vol. III., 1869. 
T have reduced my work (as formerly in my paper in the Linnean Journal, Vol. XT., page 438) to the standard of Kunth’s 
Enumeratio. Tt will be seen that I could not well have accepted Dr. Hasskarl’s numerous new genera corresponding to the old Com- 
melyna and Anilema: among other reasons because it would have necessitated my inventing another batch of new genera. he one 
new genus, which I have invented a name for, has been indicated long ago as necessary. 
T have withdrawn several of the reductions of South Indian species (especially of Cyanotis) proposed in my aboye-quoted 
paper; not that Tam_ satisfied they are wrong, but I am become very doubtful regs 
‘ding my herbarium reductions of species which 
were founded by Wight and othe 
N 
after long observation of them in the field. Ihave (since February 1870) paid a short visit to the 
s and seen many of the species alive myself: and I still incline to the views propounded in my paper of February 1870 regarding 
them; but T prefer not to reprint any opinion at present on these very difficult species. 
There is not much difference of opinion regarding the position and affinities of the order Commelynacew : they may be 
defined artificially as those lilies which have an embryo transverse to the axis (instead of in the axis) of the seed. In those Com- 
melynads which have seeds (or some of them) with a definite axis of figure (as Tab. II. figure 6, Tab. VII. figure 7, Tab, XXXIV. 
figure 3) it is easy to see by making a plane section perpendicular to the axis that the embryo lies in that plane with the growing point 
remote from the hilum or more strictly remote from that point of the hilar line in which the plane section cuts it. The growing point is 
usually not opposite the hilum but unsymmetrically turned away from it. When we come however to those genera which have small 
and altogether unsymmetrical seeds (vide Tab. XXIX. figures 3, 4; Tab. XXXII. figures 3, 4) it is only by indirect reasoning and com- 
parison with the seeds of the more typical genera that we can judge where the axis of the seed is. All we can say is, draw a line from the 
hilum of the seed to the opposite side of the seed and the growing point will be found out of that line and ina plane transverse to it. 
The artificial definition in such cases becomes difficult to work. There are a number of other more easily-seen characters 
which separate the typical Commelynacex from Liliacese but each one of these fails at some point or other. 
Thus Commelynacew usually differ by the numerous joints of their stems and the quasi-petiolate leaves: but in Anilema 
scapiflorum and other species these characters are not seen. The seeds are usually very numerous in Liliaces few in Commelynacez, 
but this distinction fades away when we come to Dichoespermum which exhibits sometimes 50 seeds to a capsule. In the typical Com- 
melyne the flowers are more or less unsymmetrical in stamens petals and ovary, but in the genera Cyanotis, Pollia, Streptolirion, 
D 
are perfectly symmetrical. Forrestia is symmetrical except that it shews in the capsule a strong tendency to the want of symmetry 
typical of Commelyna itself, and has been formerly referred to Liliaceswe, But, of these symmetrical or nearly symmetrical genera, 
Cyanotis Streptolirion and Forrestia have few seeds and those distinctly shewing the growing point transverse to the axis: while Pollia 
is so closely allied to Aclisia that it clearly stands next it. Moreover, it can be demonstrated as regards the seeds of Pollia, in spite of 
their want of symmetry, that the growing point cannot be in the axis of the seed. There remains in short no doubt, regarding any of 
the ten genera below described, that they ought to be included in Commelynacew. 
