te eat eae. ee 
=e a aoe ie Se - _ =. 
498 ROBINSON. 
well defined limits, and whether the limits assigned to them by Weddell 
and others are correct. He has been followed by Schumann and Lau- 
terbach,?” who have gone so far as to describe a species whose “Bluten 
sind leider nicht entwickelt.” The examination of Philippine material 
has shown that the problem is a very real one, and the conclusion herein 
reached is that these three genera are unusually distinct for the family 
and readily recognizable almost at a glance, but that there is a group 
whose representatives known to Weddell were by him placed in Hlatos- 
tema, though Hallier dealing with nearly allied species found that they 
traversed existing definitions and placed them in the subgenus Pellionia, 
Boerlage °** later transferring them to the genus of that name. They 
are the species included in the genus here proposed as new, with 
additions. 
Two sets of characters come into the discussion, the nature of the 
inflorescence, and that of the perianth of the pistillate flowers. . In 
typical Elatostema, both the staminate and the pistillate flowers are 
inclosed in: an involucre formed by bracts: in the simplest cases, this 
involucre consists of a single outer opposed pair, almost free from one 
another or united across their bases; with these alternate two other pairs ; 
that is to say, one of these inner bracts is situated within and nearly 
opposite the margin of each of the two outer bracts. These four inner 
bracts normally incurve at their margins and divide the receptacle into 
four parts; bracteoles surrounding the flowers or groups of flowers are 
nearly always present. In more complex receptacles, the bracts become 
more and more united, and the number may be increased, but in the 
staminate, these six bracts can always be traced, in the pistillate it 
becomes difficult to trace more than two, or sometimes even one. The 
simpler arrangement of bracts is, in general, found in species with slender- 
peduncled receptacles, and probably indicates these as the more primitive 
forms. 
In Pellionia, neither the staminate nor the pistillate flowers are in 
receptacles formed by bracts; in one of our species, there is a.tendency, 
hardly more, for the pedicels to unite to form a receptacle, much as 
happens in the majority of our species of Laportea, but this is obviously 
a very different thing from the receptacle of Hlatostema. 
Finally, in Procris, the staminate flowers are in glomerules cymosely 
arranged, while the pistillate are upon a fleshy receptacle, exinvolucrate, 
or at most with a barely projecting rim. It is evident that these three 
genera can be positively determined from the inflorescence alone, if both 
sets of flowers be available, and that Hlatostema can be distinguished by 
either alone: Procris and Pellionia have further a very distinct habit. 
The case is strengthened when the pistillate flowers are brought into 
* Nachtr. Fl. Deutsch. Schutzgeb. Siidsee (1905) 253-255. 
*8 Handl. Kenn, Fl. Ned, Ind. 3* (1900) 345, 346. 
