

OF THE GENrS CLITOBIA. 



35 



appear altogether aa we proceed westward towards the Atlantic 

 limits of Europe ; whilst the types peculiar to the extreme west 

 of Europe (excluding, of course, the Arctic flora) are wholly de- 

 ficient in America, These are among the considerations which 

 suggest an ancient continuity of territory between America and 

 Asia under a latitude, or at any rate with a climate, more meri- 

 dional than would be effected by a junction through the chains of 

 the Aleutian and Kurile Islands. 



In a systematic point of view I had formerly endeavoured to 

 render Clitoria more natural, by the elimination of DeCandoUe's 

 section Centrosema ; and I now find it necessary for the same pur- 

 pose to unite with it the N^eurocarpum of Desvaux, hitherto xmi- 

 versally adopted by other botanists, myself included. This entails 

 the giving up, as a generic character, one which, in L€guminos<je^ 

 is usually considered as absolute, the raised longitudinal nerve or 

 wing along the centre of each valve of the pod. It is the same 

 peculiarity which has induced the separation of TetragonolohuB 

 from Lotus among European plants. But in the division of LotuSy 

 as well as in that of Clitoria, this purely technical character is un- 

 accompanied by any other differences, and I have now instances 

 in Clitoria where it is inconstant in one and the same species, and 

 even on the same specimen. 



Before the introduction of Jussieu's natural system, carpolo- 

 gical characters were comparatively little attended to ; but from 

 the time he first pointed out their great importance, the absolute 

 necessity of taking them into consideration in all natural classifi- 

 cations, has been very properly insisted on by all the great bota- 

 nists of modern days. There are instances, however, in which this 

 principle may have been carried too far. The external forms ac- 

 quired by fruits in their development from the ovary to maturity, 

 and especially the foliaceous appendages they assume, are some- 

 times irrespective of their organic structure, and appear then of 

 little more consequence than the foliaceous wings or appendages . 

 on the branches, inflorescences, or calyx tubes. So also the form 

 of the membranous expansions of samaroid fruits, the consistence 

 of pericarps, the number and arrangement on the calyx and other 

 foliaceous appendages of the oleaginous deposits, called trans- 

 parent glands in Legtiminos(ey Sypericinece^ &c., or vitt<B in Urn- 

 helliferce, useful as they all may be in certain cases, as indicative 

 of general organic differences, have yet per se but little absolute 

 value in classification. This absolute reliance, in supposed confer- 



4 



mity to general principles, upon such characters, even when un- 



1)2 



J- -. 



^ _ 





