TERMINOLOGY IN MONOCOTYLBDON8. 619 



Fig. 3. The inner perianth expanded, having cast off the outer one. 



base 



tube. 



Diagrammatic Illustrations of the Genus Eriocaulon. 



Fig. 5. E. septangulare c?» showing the three outer perianth-segments much 



below the three inner small unequal ones, which alternate with the 

 three stamens, with the gland-like rudiment of the ovary in the centre. 



Fig. 6. E. septangulare § , showing the three outer perianth-segments a small 



distance below the three inner equal, erect, hyaline segments sur- 

 rounding the ovary. 



Fig. 7. E. rosulatum rj, in which the three outer perianth- segments are united 



into a spatha-like scale ; the three inner almost as in E. septangulare, 

 but larger, with six stamens. 



Fig. 8. E. rosulatum J , the outer perianth-segments distinct but reduced to two. 



Fig. 9. E. pusillum § , the perianth-segments all linear-subulate, the outer ones 



at the base, the inner at the apex of the stipes or floral axis. 



Plate VIII. 



Diagrams to illustrate the Homology of Parts in the Cyperacese. 



Fig. I. Kyllinga cylludracea, showing the spikelets disarticulated below the two 



large upper glumes, the small outer glume remaining on the base 

 of the rhachis or axis of the spikelet, which is persistent and falsely 

 described as a disk, the bract subtending the spikelet also persistent 

 and very small. Both this bract and the small glume remain in some 

 specimens persistent long after the upper part of the spikelet has 

 fallen away ; in other specimens they also fall away, leaving the base 

 of the rhachis alone persistent on the general axis of inflorescence. 

 In other species the articulation is at the base of the rhachis, and the 

 small outer glume remains on the spikelet when it falls. 



Fig. 2. Spikelet of a Cyperus, sect. Pycreus, the glumes corresponding to those 



of Graniinese (Triticeal, etc.), the two outer smaller glumes empty and 

 usually persistent after the other flowering ones have fallen off with 

 the enclosed nut. 



Fig. 3. Ficinia, showing the disk (d) f falsely described as a perigynium. 



Fig. 4. Cyperus colambetes, showing the thick white cellular portion of the peri- 

 carp (p), also falsely described as a perigynium. 



Fig. 5. Lipocarpha, the glume turned down, showing the two thin scales paral- 

 lel to the glume, and therefore possibly rather perianths than brac- 



teoles. 

 Fig. 6. Hypolytrum, two navicular bract eoles within the glume. 

 Fig. 7. Platycarpha, the two bracteoles united, and open only in an anterior 



fissure emitting the flower. 

 Fig. 8. Mapania, two free navicular bracteoles enclosing four (sometimes six) 



free flat scales, probably perianth-segments corresponding with the 



setae o f Scirp us . 

 Fig. 9. Dipla&ia, the flat hyaline scales within the bracteoles united in a flat ring 



round the ovary at the base, divided at first at the top into 4 to 6 



lobes, but soon very much torn and jagged. 



