PITTIER—-MIDDLE AMERICAN SPECIES OF LONCHOCARPUS. 41 
According to Bentham, the vexillar stamen is almost always detached 
in Derris longifolia, never in D. guyanensis+ 
GENERA ESTABLISHED WITHIN LONCHOCARPUS. 
In 1837 Vogel? founded his genus Sphinctolobium on a plant 
collected by Luschnath near Rio de Janeiro. The main characters 
consisted in a 5-toothed calyx with the teeth sometimes obsolete 
and a flat-compressed legume, sometimes 1-seeded, or 2 or 3-seeded 
and then contracted between the seeds (whence the generic name). 
These characters obviously belong to Lonchocarpus, to which Vogel’s 
three Brazilian species (S. virgilioides, S. nitidum, and S. floribundum) 
were transferred by Bentham. The special features given by Vogel 
as fundamental are not even confined to a single group, but are 
found in species widely apart. To this botanist, however, we owe 
the first mention of the fact that in Lonchocarpus the stamina] tube 
is entire, that is to say, continuously closed above the fenestrelle. 
The genus Neuroscapha, established by Tulasne* in 1843, had 
certainly a better foundation. The typeis Neuroscapha guilleminiana 
(Lonchocarpus neuroscapha) and the genus included all the species of 
Lonchocarpus in which the legume is thickened or broadened on 
the parts of the vexillar margin fronting the seeds. This is a well- 
defined feature, but as all the other characters also belong to Loncho- 
carpus, Bentham considered Neuroscapha as a simple section of this 
genus. 
In 1844 Fenzl‘ proposed, without describing it, his African genus 
Philenoptera, the principal character of which seems to have been 
the elliptic-lanceolate legume with a paper-like texture. The type 
of Philenoptera was P. kotschyana Fenzl, described by Hochstetter’ 
as P. schimperi and by Bentham as Lonchocarpus philenoptera. 
The same plant is the type also of a fourth genus, Capassa, proposed 
by Klotzsch * but never described. 
In the classification of the Middle American species given here- 
after, the name Philenopteri has been adopted to designate the first 
1 This name is not in accordance with the rules. The species was published in 1775 
in the Plantes de la Guiane, under the name of Deguelia scandens Aubl. This has 
priority over Dalbergia scandens Roxb., used for the first time in the Plants of Coro- 
mandel, volume 3, issued not earlier than 1800. The Guiana plant should conse- 
quently be called Derris scandens (Aubl.), while the East Indian species would 
become Derris timorensis (DC.), having been described in de Candolle’s Prodro- 
mus (2: 417. 1815) as Dalbergia timorensis DC. 
2 Linnaea 11: 417. 
3 Ann. Sci. Nat. II. Bot. 20: 137. 1843. 
* Flora 27: 312. 1844. 
5 In A. Rich. Tent. Fl. Abyss. 1: 232. 1847. 
6 P], Peters. Exsicc., under Capassa vielacea Klotzsch. 
