0CT . majr. 1859-60.] On the culture of Sorgho and Imphi. 299 



establishment not merely of new species but even of new varieties, 

 with the exception of one kind only derived from the seed fur- 

 nished by you to the Head Gardener. I will describe the plant 

 hereafter which appears to me to be new or at least to differ in 

 some degree from the others. 



I have, Sir, convinced myself that these supposed varieties, have 

 no existence, certainly not among the plants raised from the seed 

 received from you. They are all assignable to one type, viz., the 

 species known as the Andropogon Caffrorum of Kunth or the A. 

 Saccharatus of Roxb. to which an African origin is also assigned.** 

 The appearance and disposition of all the flowering panicles is 

 the same. In nearly all, the ramifications are verticelled or sub- 

 verticelled, the branches long, slender, flexible, subdivided into 

 numerous branchlets, or spikelets, each bearing many distinct, ses- 

 sile flowers. The lower flowers some of which are neuter, others 

 unisexual, others hermaphrodite are furnished with a single pedi- 

 celled valve ; those higher up on the spikelet have two pedicelled 

 valves representing two narrow lanceolate, acute, downy bracts 

 longer than the grains. These two glumes representing the calyx 

 are concave, beardless and entirely clothed with tomentose white 

 hairs. The stamina are as in the genus, three in number, on 

 capillary filaments, the anthers yellow inclining to light pink. 

 The ovary which is of an oval shape terminates in two longish 



* In illustration of M. Perrottet's remarks, I add the following ex- 

 tract from Kunth's Agrostographia i. p. 502. 



111. Andropogon Saccharatus, Roxb. Flor. Lid. i. 274. Erectus ; 

 paniculis verticillatis ; ramificationibus nutantibus ; calyce piloso ; 

 corollis muticis, hermaphrodita, tri, neutra univalvi. Roxb. Holcus 

 saccharatus Lin. Sp>ec. 1484 — Willd. Spec, 4-930. Sorghum sacchara- 



tuna, Pers. Syn. l'10l=Host. Gram. 4 t. 4 — Holcus Dochna [ar^~LjJ 



Yorsk. JEgyp. 174, (teste Delile) — India orientalis, Arabia,©. An 

 planta Roxburghiana ccrte eadem ac Linneana ? 



112. A. Caffrorum, Kunth Gram. 165. Glumis villosis ; seminibus 

 compressis inermibus. Thunb. — Holcus Caffrorum, Thurib. Prod. 20. 

 Ejusdem Flor. Cap. i. 410 — Willd. Spec. 4*930 — Holcus cafer, Arduin, 

 Saggi di Padov.— 119 t. l.f. 1. — Sorghum Caffrorum, Beauv, Agrost. 131, 

 Sorghum Arduini, Jacq. Eel. Gram. 25. t. 18. Cap. Bon. Spec© 

 Sprengel cum precedente conjungit. 



Vol. xx. o. s. Vol. vii. n. s. 



