BENTHAM ON PAriLlOK ACE;E AND C^SALPINIEAI. 125 



^U-~ Observations on the Distinctive Characters of the 

 Papilionace^ and C^salpinie^-, Sub-orders of Legumi- 

 NosiE. By George Bentham, F.L.S, 



My friend Dr Vogel, who has for some years past studied 

 with great care the Order of Legumitwsce, and published 

 several important memoirs on the subject, has communicated 

 to me a paper on the plants of that Order collected by 

 Meyer, in his voyage round the world.* To this paper he 

 has prefixed some criticisms on the limits I had proposed to 

 draw between the sub-Orders Papilionacece and CtesalpiniecB, 

 which have induced me to repeat some of the investigations 

 I had gone into, and to give the matter further consideration, 

 the results of which it is now my object to state. 



Dr Vogel's remarks are founded on the opinions emitted 

 in a memoir I prepared at Vienna in the commencement of 

 the year 1837,-f- and on two short papers read before the 

 Linnaean Society, one on the genus Mora, read March 20th, 

 1838;]: the other on J rachis and Voandzeia^ read May 1st, 

 1838.§ At the time I published these partial memoirs, I had 

 examined but few C<ssalpinie(t', and although it even then 

 appeared to me that the structure of the flower would furnish 

 the best character, I had not formed a sufficiently definite 

 notion of what really constitutes a papilionaceous corolla, to 

 make use of it as a positive character, and had been led into 



* From the Nov. Act. Acad. Cfps. Leop. Carol. Nat. Cur. v. xix. 

 Suppl. 



t Published in the Annalen des Weiner Museums der Naturgeschichtc, 

 V. xi. p. 63. et seq. 



t Linnaean Transactions, v. xviii. p. 209. 



§ Linngean Transactions, v. xviii. p. 155. Neither Dr Vojel nor Dr 

 Walpers appears to have read this paper through when thev quote it, for 

 the former says that I refer in it, Arachis and Voandzeia to Hedysarea, 

 and Dr Walpers (Linna;a v. xiii. p. 531) quotes it as his authority for 

 placing Voandzeice among Hedysarece, when one distinct object of the 

 paper was to show, that these two genera were but slightly related, and 

 that while Arachis should be placed among Hedysarece, Voandzeia belongs 

 to PhaseolecB. 



