518 



MR. G. BENTHAM ON CLASSIFICATION AND 



received a separate name ; that of lodicules is the one generally 

 adopted, and does not appear objectionable, although the classical 

 meaning of the word may not be very applicable. 



To sum up, therefore, — the spikelets of Oraminese may be de- 

 scribed as composed of a series of alternate glumes, distichously 

 imbricated along the axis, one or two (rarely more) at the base of 

 the spikelets and sometimes a few at the end empty, the others 

 having each a single sessile flower in their axil, the short floral 

 axis bearing a palea, two or rarely three lodicules, and three (rarely 

 two) or six stamens under the terminal ovary. 



In conclusion, I have only to add that in making these observa- 

 tions I have had nothing new to bring forward, my only object being 

 to enforce a principle generally admitted, but unfortunately too 

 much neglected by speculative botanists, viz. that to be really 

 useful, descriptions should, in clear and intelligible language, not 

 only enable the reader to identify the plant he has in hand, but call 

 his attention specially to those characters which may indicate its 

 real affinities, the homologies of its parts, and any other relations 

 they may have. But for this purpose it is necessary that the 

 author should distinguish descriptions of plants from theoretical 

 explanations, that he should, in terms the most capable of strict 

 definition, describe only what the observer may actually see, not 

 what it may be theoretically imagined he ought to see, reserving 

 his theories for comments upon what has been actually observed. 



P.S. Since writing the above, a further detailed examination of 

 Lipocarpha has tended to prove that the two scales within the 

 glume are the homologues of the hypogynous scales of Fuirena 

 and of the flat scale-like setae of Scirpus littoralis — these scales 



mg one m Remicarpha, two in Lipocarpha, three in Fuirena, 

 four in Scirpus littoralis, and deficient in Scirpus sect. Isolepis, 

 genera all very closely allied to each other.— a. B. 



DESCRIPTION OF TIIE PLATES. 



Plate VII. 



_ p 



Diagrammatic Illustrations of the Genus Xyris. 



Fig. 1. Young bud showing the two bracteoles (66) enclosing the outer perianth 



{c) y which then completely envelops the inner perianth, of which the 

 tube has not yet grown out. 



Fig. 2. The same further advanced : the inner perianth in lengthening has 



forced up the outer one, which has left a scar where it bad been 

 attached, and i9 opening on one side to enable it to be east off. 



