460 APPENDIX. No. V. 



of CjT5erace£e, and in Lipocarpha itself, from which, however, they are still 

 sufficiently different in their relation to the including squamae and to the axis 

 of the spike. 



This view of the structure of Hypaelytrum, of which there is one species in 

 the Congo herbarium, appears to me in some degree confirmed by a comparison 

 with that of Cliondrachne and Clvorizandra ;* for in both of these genera the 

 lower squamse of the ultimate spikelet are not barren, but monandrous, the 

 central or terminating flower only being hermaphrodite. 



GRAMINEiE. Of this extensive family there are forty-five species from 

 the Congo, or one twelfth of the Phfpnogamous plants of the collection. This 

 is very nearly the equinoctial proportion of the order as given by Baron 

 Humboldt, namely, one to fifteen, with which that of India seems to agree. 

 On the north coast of New Holland, the proportion is still greater than that 

 of Congo. 



The two principal tribes which form the far greater part of Gramineae, 

 namely, Poacece and Panicew have, as I have formerly stated,-f- very different 

 relations to climate, the maximum both in the absolute and relative number of 

 species of Paniceae being evidently within the tropics, that of Poaceae beyond 

 them. 



I have hitherto found this superiority of Paniceaj to Poaceas, at or near the 

 level of the sea within the tropics, so constant, that I am inclined to consult 

 the relative numbers of these two tribes, in determining whether the greater 

 part of any intratropical Flora belongs to level tracts, or to regions of such 

 elevation as would materially affect the proportions of the principal natural 

 famiUes : and in applying this test to Baron Humboldt's collection, it is found 

 to partake somewhat of an extratropical chai"acter, Poaceae being rather more 

 numerous than Paniceae. While in conformity to the usual equinoctial propor- 

 tion, considerably more than half the Grasses in the Congo herbarium consist 

 of Paniceae. 



Among the Paniceae of the collection, there are two unpubhshed genera. 

 The ^rst is intermediate, in character, to Andropogon and Saccharum, but 

 with a habit very different from both. The second, which is common to other 

 * Prodr. Flor. Nov. Hott. l,p. 220. 

 t Prodr. Fl. Nov. Holl, I, p. 169. Obs. I J. Flinders' s Voy. 2, p. 583. 



