OF THE GENUS MILLETTIA. 137 
new genera in order to restore it to manageable systematic limits. Thus 
Craibia * was established to accommodate a group of African species with 
alternate leaflets, Adinobotrys t certain Indo-Malayan trees and climbers with 
indehiscent pods. Padhruggea f was reinstated for a somewhat similar group. 
Fordia § was enlarged to receive a number of shrabs with 2-ovuled pistils. 
These investigations also led to the establishment of Leptoderris || and 
Ostryoderris T and to the extension of Ostryocarpus **, not because these 
genera were nearly allied to Millettia, from which their Derris-like fruit at 
once distinguished them, but because, in the absence of fruit, many species 
properly belonging to them bad been included in Millettia and had in this 
revision to be placed elsewhere. 
A large number of Millettia species have been given names under Phaseo- 
lodes, Mill. by Otto Kuntze tf, and under Araunhia, Raf. by Prain 1i 
(Indian spp.) and by Greene §§ (chiefly Australian spp.), on the assumption 
of their congeneric relations with Wistaria, which is antedated by the two 
latter. The view adopted by the writer that Millettia and Wistaria are 
distinct in consequence of their inflorescence renders the citation of the 
synonymy thus introduced unnecessary in this paper, but the Avaunhia 
and Phaseolodes combinations already published may be found by reference 
to the books mentioned. 
МҮЗТЕМАТІС POSITION. 
The Leguminose which have woody stems, unequally pinnate leaves, 
papilionaceous flowers, and coherent stamens were collected by Bentham 
and Hooker in their ‘Genera Plantarum’ into two tribes, Galegee and 
Dalbergiee, according to the dehiscence or indehiscence of their pods; and 
although there are certain genera possessing the above characters which 
cannot for other reasons be included and vice versd, botanists are generally 
agreed that these tribes represent a natural division of the family. In the 
same way, the fact that some genera of Galegec possess pods which do not 
open at maturity, but are nevertheless otherwise clearly referable to that 
tribe, does not affect the usefulness of Bentham’s diagnostic character. 
The Galegew were again divided into 7 sub-tribes, 5 of which were based 
principally upon special single characters, leaving Tephrosiee and Robiniew 
to divide between them the typical elements of the tribe. They are not 
in fact, themselves very readily divided and are distinguishable rather by 
tendencies than by characters. It will be sufficient, however, here to note 
* Dunn in Journ. Bot. xlix. (1911) 106. q Dunn in Kew Bull. 1911, 363. 
+ Dunn in Kew Bull. 1911, 194. ** Dunn, l. c. 362. 
і Dunn, 4. c, 197. tt Kuntze, Rev. Gen. Pl. 1. 201. 
$ Dunn, } c. 69. tt Prain in Journ, As. Soc. Beng. lxvi. п. (1897) 358. 
|| Dunn, /. c. 1910, 386. $$ Greene in ‘ Pittonia, ii. 175. 
