of the Monandrous Plants of India. 273 
added two other species, dichotomum and virgatum; but I am 
inclined to believe that both these plants, if not already known 
to us, will be found on investigation to belong to other genera. 
Dr. Roxburgh, indeed, admits that the habits of his three species 
of Phrynium are different, although he conceives they agree in 
their generic character; but I have commonly found the true 
generic distinctions confirmed by the habit of the plant, and 
am doubtful when this is not the case. 
"That the three genera of Maranta, Thalia, and Phrynium are 
nearly allied. to each other, is I think evident, as appears more 
particularly by the seed, in which the albumen of the nuciform 
fruit is pierced by the thread-like embryo; yet their generic 
distinctions, as well as their habits, seem to require their separa- 
tion. In Maranta the anthera is irregularly placed on the mar- 
gin of the petal or petal-like filament: sometimes on the right, 
and at others on the left; but the edge, where the anthera is - 
found, is always thickened downwards, as if by a concealed sta- 
men; and in some instances this stamen is even separated trom 
the petal, for a very short distance, immediately below the an- 
thera. . In Thalia the anthera is placed in the middle of its pro- 
per filament, opposite to which is the short style, terminating in : 
an irregular ringent stigma, resembling the mouth ofa beaker, ` 
and. wholly different from that of any other genus in the whole 
order. In Phrynium, the anthera is placed in front of a strong 
arched or inflexed filament, in such a- manner, that if it were 
erect, the anthera would appear to be attached to the back, 
whilst the stigma is simply funnel-shaped, in which it agrees with 
the chief part of the scitaminean tribe. "These distinctions, in 
themselves so important, are confirmed by the respective habits of 
the plants ; that of Maranta being ramose, and frequently dicho- 
tomous; Thalia flowering terminally on a long stem from the 
centre 
