Graminece. 115 



which may be regarded as genera, these as such do not disturb the 

 classification by being referable to other tribes. 



As stated in the Preface to Part IV., no materials were left in MS. 

 by Dr. Trimen for the elaboration of the grasses, my resources are hence 

 limited as regards habitats, colours, and flowering seasons. For the 

 former I am dependent on those given in Thwaites's ' Enumeratio ' in 

 the Peradeniya Herbarium, where unfortunately many species have no 

 habitat or date of collection assigned to them, and especially in the late 

 Mr. Ferguson's valuable essays on the grasses of Ceylon.* I am also 

 indebted to Mr. H. H. W. Pearson for the loan of his carefully ticketed 

 collection of grasses, made chiefly in the patanas of the island in 1898. 

 For the flowering seasons of the grasses I am altogether deficient of 

 resources. As with the Cyfteraceiz, they would seem, from such dates as 

 I find attached to some specimens, to be either greatly prolonged, or to 

 occur twice in the year, dependent on the monsoons, the wet and dry 

 climates, elevation above the sea, and exposures. A full account of the 

 distribution and flowering seasons of the Ceylon grasses could not fail to 

 be both instructive and interesting. The two charts issued with this 

 Part of the Handbook, of the rainfall and forest regions of the island, 

 indicate data as to climate and geographical areas, available for an 

 essay on this subject. 



Series I. PanicaceSB- — Spikelets usually articulate on 

 their pedicels and deciduous from them when in fr., rarely- 

 persistent and deciduous with them, binate, a sessile fruiting 

 and a pedicelled male or neuter, rarely solitary; glumes 3 or 4, 

 rarely only 2, rhachilla of spikelets rarely produced between 

 the bases of the glumes or beyond the upper glume ; sessile 

 spikelets 1-2-fld. (never more), upper fl. alone (if 2) fruiting 

 — Herbs, Spinifex alone shrubby of Ceylon genera. 



Exceptions. — Spikelets persistent or tardily deciduous, or glumes I 

 and II separately deciduous in Isackne, Chamceraphis, Pennisetum, 

 Stenotapkrum, Thuarea, Arundinella, and Zoysia. Spikelets inarticulate 

 at the base and deciduous with the internodes of a fragile rhachis in 

 many Andropogonece. Both fls. of the spikelet of some species of 

 Isachne ripen fruit. 



Spikelets articulate on their pedicels and deciduous occur in Poacece, 

 in Garnotia, Polypogon, Sporobolus, and Lophatherum. Lower fl. in the 

 spikelet male and upper fertile, occur in Phragmites, &c. 



A. Rhachis or branches of infl. inarticulate ; pseudo-articulate in 

 Stenotaphrum, articulate in Trachys. Some genera of Tribe Andro- 

 pogonecB have inarticulate branches, racemes, or spikes. 



* ' Notes on the Grasses and their Distribution in Ceylon, by W. 

 Ferguson, F.L.S., read before the Ceylon Branch of the Royai Asiatic 

 Society (1880).' I have seen no copy of this paper, only a precis of its 

 contents kindly made for me by Mr. Lewis of the Forest department. 

 'Grammes or Grasses Indigenous to or Growing in Ceylon, with Notes 

 especially on those used as Fodder Plants, by W. Ferguson, F.L.S. 

 Colombo, 1886.' I have made much use of this instructive work, though 

 perhaps not all that I might have, had I seen corresponding specimens 

 that would have certified the nomenclature of some of the critical 

 species. Mr. Ferguson's list embraces 218 species, including those 

 known only in cultivation. 



