46 T. Holm — Studies in the GyperacecB. 



nowii, where the beak is shorter and less pronounced. How- 

 ever the internal structure is the same for these two species. 

 The outer wall of epidermis on the dorsal face is somewhat 

 thickened, on the ventral face it is much less so ; there is but little 

 mesophyll displayed in a few strata and no stereome, not even by 

 the two mestome-bundles, in which, furthermore, the mestome- 

 sheath is quite thin-walled. In G. Backii the utricle shows a sim- 

 ilar structure with the exception of a few stereome-cells on the 

 leptome-side of the two mestome-bundles. A much firmer 

 structure is exhibited by G. oedipostyla in which the epidermis 

 is quite thick-walled on either face, besides that the mesophyll 

 is more compact and occurs in several layers ; moreover we 

 find in this species very thick-walled stereome, not only on 

 either face of the mestome-bundles, but also as several isolated 

 groups between these. There are more than two mestome- 

 bundles and these possess a very thick- walled mestome-sheath. 

 The utricle of G Geyeri and (J. midticaulis has a very short 

 beak like C. cedipostyla y it is perfectly glabrous and smooth. 

 The epidermis is heavily thickened on both faces ; the meso- 

 phyll is poorly developed, and the stereome, which is quite 

 thick- walled, is in G Geyeri confined to the mestome-bundles, 

 while in the other species it occurs as several isolated groups 

 between them. 



It would seem from the above as if C. Steudelii, C. Willde- 

 nowii and G. Backii are closely related to each -other, but not 

 to C. cedipostyla, C. Geyeri or G. midticaulis, hence the sec- 

 tion Phyllostachys, as defined by recent authors, does not con- 

 stitute a natural section. If we, moreover, consider those 

 species, which by Torrey were the fundamental ones for his 

 genus, it appears as if they are not sufficiently characteristic 

 among themselves to necessitate the establishment of even a 

 section. And the same is the case with G. cedipostyla. In 

 regard to G. Geyeri and C. midticaulis, these do not possess 

 very pronounced characters either, but similarly to the others 

 may be arranged in some of the larger sections as " formse 

 hebetatse." The species of Torrey's Phyllostachys appear, on 

 the other hand, as inseparable from the Hymenochlcenw, of 

 which they probably represent one of the earlier types. The 

 earliest is unknown in this section, but was perhaps mono- 

 stachyous ; but we might, also, suppose that the Hymenochlcence 

 developed parallel with one of the other sections from one 

 common " forma hebetata," which then branched out into two 

 or more distinct sections. 

 Brookland, D. C, February, 1900. 



