510 ME. G. BENTHAM ON CLASSIFICATION AND 



below and more or less round the seed-bearing cavity, observable 

 in the ripe caryopsis of several Cyperaceae belonging to various 

 genera. This genus Anoporwm, originally founded by Nees on 

 the Oyperus cephalotes, has, under Boeckeler's revision, become 

 an unnatural amalgamation of four or five Cyperi taken from very 

 distinct sections of the genus, with Scirpus cubensis, Kunth, to 

 which has been added as a synonym the Oxycaryum SchomburgJc- 

 ianum, Nees, a Guiana plant resembling it in aspect but with 

 differently shaped glumes and without that thickening of the 

 pericarp supposed to be characteristic of Anoporum. G-risebach 

 has recently, and perhaps correctly, reduced Oxycaryum to a section 

 of Scirpus ; and the other Anopora must be restored to Gyperus. 

 A still less intelligible misnomer appears in the descriptive 

 works of some German and Italian botanists, who give the name 

 of perigonium (usually taken as a synonym of perianth) to the 

 hyaline wings which border the angles of the rhachis in the spike- 

 lets of some Cyperi. These wings are decurrent from the margins 

 of the glume next above ; and though they sometimes become de- 

 tached in the shape of small scales, they have no connexion what- 

 ever with the flower of which they are represented as the perigonial 

 scales. 



In some Cyperaceae, within the bracteoles or independently of 

 them, and immediately under the slamens or alternating with 

 them, or even apparently almost within them, are six or fewer, 

 rarely more than six, setae or narrow usually hyaline scales. These 

 are very generally supposed to represent perianth-segments ; and 

 there is every reason to believe that that is their real homology. 

 Yet this is scarcely yet absolutely proved, and it may be better 

 to continue to describe them by the general terms of hypogynous 

 setae or scales, which commit the author to no theory. Amongst 

 the genera which have no bracteoles, it is in Fuirena that they 

 come the nearest in appearance to true perianths. They there 

 consist usually of three short, broad, hyaline scales, alternating 

 sometimes with three setae. In Scirpus and several of its imme- 

 diate allies, they are usually all reduced to setae, six being appa- 

 rently still the normal number ; but they are often reduced in 

 number or very unequal and irregular, exceedingly deciduous or 

 entirely wanting, and even varying in these respects in the same 

 species, on which account Boeckeler and others have reunited Iso- 

 lepis with Scirpus. In a few species of Scirpus, Rhynchospora, etc. 

 the number is often slightly increased, and in some species oiErio- 



