PHILIPPINE URTICACEAE. 499 
the discussion. In the species here placed in Hlatostema (excepting, 
of course, those whose pistillate flowers are unknown), there is always 
a pedicel, rarely as much as 1 mm long but usually shorter, at the apex 
of which there is a cup-shaped perianth, with very short, rounded lobes, 
sometimes ciliate, but this last is a most inconstant character. Ap- 
parently these are treated by Stapf °* as staminodes, but this interpreta- 
tion would involve the presence of staminodes at the base of an ovary 
raised upon a common pedicel, which is quite distinct from anything else 
in the family. So far as appearance goes, they might well be staminodes, 
the lobes are in most cases about 0.1 mm long, and number 3, they are 
often so little defined that it is difficult to be certain of the number, 
which may appear to be 2 or 4: in one case it was definitely 4 in flowers 
from a receptacle in which others were as definitely 3. This perianth 
is always present, and usually merely invests the extreme base of the 
ovary, rarely it reaches one-third or even one-half of the length of the 
latter; in two species, flowers were found whose perianth appeared quite 
typical, but on closer inspection, it proved that the lobs were infolded 
and that when outspread they were nearly as long as the achene, but this 
was of so little importance that other flowers in the same receptacles 
were quite typical. Staminodes, in such flowers, must be conspicuous 
unless extremely minute: I have probably not found them in any species, 
though in a very few cases minutes spots were seen adhering to the base 
of the ovary: in F#. stracheyanum, Weddell describes them as longer 
than the perianth. 
In Pellionia, the pistillate perianth is 4- or 5-parted, usually about 
as long as the ovary but sometimes distinctly shorter, and its divisions 
are further terminated by a spur of similar length. This spur varies 
slightly in its position, even in a single flower, being usually distinctly 
dorsal just below the apex of the wider sepals, or separated from the 
apex by a minute rim, or quite terminal: it is superficially the most 
conspicuous thing in the flower. ; 
In Procris, the perianth is so deeply parted and widely spreading that 
it becomes mechanically very difficult to allocate its divisions to the 
corresponding ovary or achene; these divisions are 3 or 4, as long as the 
ovary or considerably shorter than it, and always lack the spur of 
Pellionia. They can not be mistaken for those of either that genus or 
still less of Zlatostema, indeed they more nearly resemble those of some 
species of Pilea. Staminodes are probably wanting. 
If this were all, no generic difficulties should arise. But Weddell 
included in Elatostema a very few species, of which 2. manillense Wedd. 
may be taken as typical, in which the inflorescence is often very con- 
densed but not contained in a bract-formed receptacle, where the pistillate 
perianth is 5-parted, at the most very shortly acuminate, and where short 
7? Trans. Linn. Soc. Bot. IT 4 (1894) 231. 
