I—cOo} 
[ELYNA. Kunth Enum. IY., p. 35 (Char. emend.) 
SSeS 
Stamina 3 antica perfecta (quorum intermedium dissimile) 3 vel 2 sterilia difformia. Capsula 3-aut 2-locularis ; 
5-4-3-aut 2-ovulata. Pedunculi 2, (quorum unus haud raro rudimentarius vel obsoletus) ex spatha complicata aut cucul- 
lata orti eaque subinclusi, 
Sect 1—Capsularum loculi duo 2-ovulati: loculus tertius 1-ovulatus vel 0. Semina normaliter 5 vel 4. 
a. Eu-commelyna (= Commelyna genus Hasskarlii) Loculus tertius evolutus. 
b. Dissecocarpus (genus Hasskarlii) Loculus tertius deficiens. 
Sect 2.-Capsularum loculi 3 aut 2omnes l-ovulati. Semina 3 aut 2. (= genera Trithyrocarpus et 
Spathodithyros Hasskarlii. 
Dr. Hasskarl’s Monograph of the Commelynaces of Malaya and my paper on the Commelynacew of Bengal were prepared 
independently ; and it is satisfactory that we have both been led to establish our main distinctions on the capsules and seeds. Our work 
differs widely at first sight, because I have made sections where Dr. Hasskarl has made genera : indeed, some of his genera appear to me 
untenable even as sectional divisions. 
Inmy paper I formerly based my main sub-divisions of the genus Commelyna (mihi) upon the dehiscence of the capsule. In 
many species one cell is very constantly indehiscent and withers so closely on the seed that it is with difficulty removed therefrom by 
scraping with a knife ; the other two cells being early dehiscent, Vid. Tab. I. figure 4,, Tab, TV. figures 4, 5. In other species all the cells of 
the capsule dehisce. Vid, Tab. IX. figures 3, 10. But this distinction I find is not absolute. In Commelyna communis for instance one 
seed is nearly always retained in a closed cell, but I have found plants grown in very wet places in the hills where the closed cell dehisces 
by as clean-cut a loculicidal line as the other cells, but much later. Hence I have abandoned the differences in dehiscence as the main 
ground of sub-division of the genus. 
The number of cells in the capsule is not constant in some species. In Commelyna app2ndiculata the third small closed 
cell is usually present but sometimes quite obsolete, so that both 3-and 2-celled capsules are found on the same plant. Vid. Tab. XIII. 
figures 1, 2, 3, 4, 8,9, 10. In Commelyna obliqua the cells are generally 3 sometimes 2 ; and I find both 2-and 3-celled capsules on the 
same plant. Vid. Tab. IX. figures 1, 2,3, 8,9, 10. In Commelyna salicifolia I have found in some Assam specimens 2-and 3-celled 
capsules on the same plant. ‘The genora Trithyrocarpus and Spathodithyros of Hasskarl on these facts can har dly be maintained. 
The number of barren aborted stamens is in my experience always 3 or 2 in Commelyna, but I should be sorry to found a 
new genus on a Commelyna because it exhibited less than 2 barren stamens. Spathodithyros is defined by Hasskarl (in his Commelynacez, 
p. 10) as having no sterile stamens ; but the single plant of the genus, Commolyna suffruticosa, is described by Hasskarl (in Plante Jung- 
* JT find 8 barren stamens. 
huhnian, p. 141) as having “ stamina generis.’ 
Heterocarpus, Wight, has not been found in Bengal. I hazarded a suspicion (in my paper on the Commelynacew of Bengal) 
that my Commelyna rajmahlensis was Heterocarpus glaber of Wight : two of the seeds early disperse together with the capsule valve, and 
there remains one seed permanently held in the single loculus that withers upon it. 
As to Heterocarpus obliquus (Hasskarl Commelynacex, p. 7,) the author must have had very imperfect specimens to have 
so placed it, as it is the type of his genus Trithyrocarpus. I consider Heterocarpus ? obliquus Hassk. and Trithyrocarpus paleatus 
Hassk. to be one and the same plant. My Javan specimens however are not authenticated by Dr. Hasskarl himself, but issued named 
from the Buitenzorg Herbarium. It only requires in my opinion to read Dr. Hasskarl’s exposition of Heterocarpus Wight (in his Com- 
melynacew, p. 6,) to be satisfied that the genus is bad. 
In Commelyna, so far as I know, those species in which some of the cells of the ovary contain two ovules never vary so 
that all the cells of the ovary contain one ovule only : that is to say my two main sections now proposed of the old genus Commelyna 
are strictly definite. Nevertheless, after considering the variations which occur in Forrestia Hookeri and Cyanotis nobilis, I could not 
feel altogether surprised should my Commelyna sikkimensis turn outa variety propagatable from Commelyna communis. Should how- 
ever this be found to be the case, I fear all the new genera carved by Dr. Hasskarl out of the old Commelyna will be swept away as well as 
my ceaidionll divisions. 
The markings on the seeds of Commelyna are very trustworthy. The arrangement of the spathes (i. e. scattered axillary 
or crowded towards the end of the branches) is also very characteristic in some species. 
As to the shape of the spathes, it varies enormously on individual plants, sometimes even more widely than is shewn in my 
picture of Commelyna communis, Tab. I. 
The peduncles are normally two in each spathe, whereof the lower comes into flower first: this lower peduncle rarely carries 
more than two flowers, the upper peduncle carries from two to twenty, according to the species. Hach flower is pedicellate and all the 
pedicels are articulated on the peduncle: such flowers as are barren or wholly male usually drop early by the disarticulation of the pedicel : 
where seeds are perfected, the disarticulation never takes place. The lowermost i. e. first-flowering flower of the upper pedicel is always 
perfect, and on this upper pedicel there may be in some species 1 to 6 flowers producing capsules, the tendency of each successive higher flower 
being more and more to become male. In the lower peduncle the flowers are rarely both perfect ; in some species not infrequently one is per- 
