152 
with a thick, coriaceous epicarp splitting away at maturity from 
the membranous endocarp, which remains investing the seeds. 
Daubentonia has a several-seeded, compressed and often torulose 
legume, the central portion of the winged vaives becoming indur- 
ated over the seeds, and hence never dehiscent.* A more striking 
contrast of characters in a group of nearly related genera it would 
be hard to find elsewhere in the Leguminosae. 
The bibliography of the species is involved in some obscurity, 
owing to the fact that all copies in this country of one of the works 
concerned, Ortega’s Hortus Matritensis, are apparently incomplete, 
and to the resultant fact that the synonymy has been thrown into 
confusion by later writers. The genus Daudentonia was based by 
De Candolle upon AL schynomene longifolia Cav.,and Piscidia punicea 
Cav. Reference to the latter’s work shows the plate of A. /ongz- 
Solia to be an excellent and unmistakable reproduction of our Tex- 
ano-Mexican plant. 
In addition to the two species of Daudbentonia, De Candolle de- 
scribes in Prodromus 2: 265 a species of Sesbania under the name 
of S. longifolia, for which he cites as a synonym Ortega’s A Eschy- 
nomene longifolia, referring to the plate of the latter in the ninth 
decade of the Hortus. In his characterization of the species he 
uses the words “‘ leguminibus linearibus torulosis acutis,” which can- 
not apply to anything but a true Sesdamia. It is thus evident that 
De Candolle understood with great clearness the fact that Ortega 
and Cavanilles had described two different plants under the same 
binomial appellation (A Lschynomene longifolia), the former of which 
he transferred to Sesbanza, and the latter he raised to generic rank; 
by this means he preserved the same specific for each. It is due 
to the efforts of later botanists to combine both forms under Ses- 
éania that most of the trouble has originated. Among British 
authorities, especially, there has been a surprising tendency to 
jump at conclusions, and, as it would seem, to avoid the proper 
verifications of referemces, since a very superficial examination of 
the pages abo would have affordeda solution. 
The genu: tonia is taken up by Torrey and Gray, and 
the citations are correctly given.t On the other hand, Hemsley, 
2 * For a good figure of this type of legume, see Fig, 117, Engl. & Prantl. Nat. 
es 47% 1894, wader Seshawis punicea. 
Bixee Sencn* Am. 1: 294. oP. a 
