116 MR. BENTHAM ON MALVACE® 
with the anthers in fours between each two sterile stamens ; T. uni- 
flora, DC., and T. parviflora, Boj., with the anthers in threes ; and 
T. decanthera (Melhania decanthera, DC.), T. laurifolia (Melha- 
nia laurifolia, Boj.), T. erythroxylon (Melhania ergthrozylon, Ait.), 
and T. melanoxylon (Melhania melanoxylon, Ait.), all with two 
anthers only between each two sterile ones. All the species are 
from Mauritius or Madagascar, except the two last, which are 
from St. Helena—or rather were, for both are now said to be ex- 
tinet. This distribution of so marked a genus over these distant 
islands, without any traces of it (as far as known) in the interme- 
diate continent, may suggest some curious speculations as to the 
gradual extinction of ancient floras. These two St. Helena spe- 
cies are indeed described as pentandrous only ; but I have certainly 
found the anthers in pairs in all the specimens I have seen, 
although with their short filaments united: that is, however, par- 
tially the case in some of the Mauritius species. 
PENTAPETES, Linn. 
This genus, occasionally made the receptacle of several doubtful 
Dombeyez, is now reduced to the single P. phenicea, Linn. ; for 
the P. angustifolia, BL., is generally admitted to be a mere variety. 
Miquel has indeed distinguished it as a genus under the name of 
Eriorhaphe; but on carefully studying his description, I find 
every part of it (including the nerve-like plumose placenta, whence 
he derived his name) applicable to the common P. phenicea, ex- 
cept, perhaps, the number of anthers, 10 only instead of 15—that 
is, two instead of three between each two sterile stamens. From 
having observed, however, that one or two anthers are wanting in 
some flowers of our specimens, I should suspect that the number 
10 was accidental in the flower examined by Miquel. 
Mernanta, Forsk. 
Melhania, deprived of the arborescent species referred as above 
to Trochetia, and including Brotera, Cav., Pentaglottis, Wall., 
Cardiostegia, Presl, and Vialia, Vis. becomes a very natural 
and well-defined genus, distinguished from all other Dombeyee 
by the anthers solitary between each two sterile stamens, and 
readily known by their habit approaching that of Serrea among 
Malvacee rather more than that of Hermannia, to which it has 
been compared. It includes about sixteen species, dispersed over 
Africa and the warmer, drier regions of Southern Asia and North- 
ern Australia. 
