196 MR. TOKUTARO ITO ON A NEW 
anthers, each bilocular, which proves their affinity to the Balano- 
phora dioica, Wall., group. Now, besides B. dioica, there occurs 
a nearly allied species known as B. polyandra; and the difference 
between the two species is considered to be scarcely recognizable, 
since so competent an observer as Sir J. D. Hooker * says:—“I `’ 
have frequently not been able to distinguish female specimens of 
this (B. polyandra) from those of B. dioica.” The difficulty of 
distinguishing the two species does not seem to depend entirely 
on their close resemblances, but also on their variability, B. dioica 
having been described by Prof. Eichler t as “ planta magnitudine 
variabilis." The distinction between the two species, however, is 
founded entirely on the diversity of the stamens, namely the 
anthers in those of B. dioica are bilocular, while in B. polyandra 
they are multilocular. Depending principally upon the above 
determinations, in an examination of the male flowers of Riukiu 
specimens, perhaps we may not be far wrong in presuming that 
the plants we found in Japan are probably B. dioica, Wall. The 
only difficulties which I cannot pass over without mention, are, 
that in our Riukiu specimens the number of the bracts of the 
peduncle is rather fewer, and they are loosely placed and irregu- 
larly imbricate as we generally observe them to be in B. poly- 
andra, while in B. dioica the imbrication is remarkably regular; 
and that the capitulum of the Riukiu specimen approximates more 
in shape to B. polyandra than to B. dioica. By comparing our 
plant with other species, it differs essentially from B. elongata by 
its tubular rhizome, the latter having it elongate, from .B. Har- 
landi by its non-globose capitulum, and from B. involucrata in 
other respecta. 
With regard to the distribution of B. dioica, I have been able 
to examine the following specimens, all of which are contained in 
the Herbarium at Kew :— 
India: Sikkim, alt. 4000--7000 ped. (Sir J. D. Hooker, d et 2); 
Khasia, alt. 3000-5000 ped. (Sir J. D. Hooker, d et 9, Dr. 
Wallich 1849, 3); Nepalia (Herb. East Ind. Comp., d , 1821). 
The allied species, B. polyandra, is represented from various 
localities, some of which may be given thus :— 
India: Himalaya orient. (Herb. Griffith, n. 2435, g); Sikkim, 
alt. 5000-7000 ped. (coll. Sir J. D. Hooker, d et 2). 
* Trans. Linn. Soc., Bot. vol. xxii. 1859, p. 47. 
t De Candolle's * Prodromus, xviii. pp. 145 & 149, 
