108 O- King — Materials for a Flora rtf the Malayan Peninsula, [Ifo. 1, 
Pekak ; Lamt, Goping, etc. Kunstler 4028 ! 4149 I 4504 ! 8551 ! 
Scorfechhd 1101 Malacca; GriJJith 1774! Singapork ; Ridlmj ! Ujstrib* 
Teuasserim ; Borneo. 
Var. ? aptera; pod 1 in. wide, wingless; leaflets with lateral veins 
stronger beneath. * 
Malacca; Maingay 613 S Pbkakj Kumtler 4518 f 6428! 
Vab. ? miUettioides pod and Icafli^ta as in vak. ? aptera but the 
former nsiaallj loager and ultimately dehisciug (as in Millettia) along 
both satarea. 
Perak ; Ulii Biibong, Kimsfler 10G96 f 
The plant liere deacribod as Deri'iH malaccensis ia extremely closely related to 
JO. cmieijolia of whioh indeed it was treated by Mr. Bentham as n variety. Its 
leaves differ mainly in having fowor bnt Inrger IcafletB with \oitg caiidate-actiminatQ 
tipa ; the flowers, too, are considerably larger and of n somewlmt different colour : 
the pods o£ D. malaccensis are also much larger than those of D, cuncifolia. It 
ronst also, from the deacriptioti of that plant, be very nearly iiUied to D. nimitana 
Beuth- {PL Juntfh, a Java species not represented ia ITerb, Calcntta. Tho 
foHagio of tho two is evidently almost identical but the Rowers a lilllo hirger in 
mon faint, he'm^ '75 in. long. D. malaccentiis is tbns evid^'jitly jnteimediate be- 
tween D. cuneifolia and D. montatia as regards its petals; it in likewise intemiodiato 
as regjurda ovary, Mr. Bentham asoribes two ovnles to D. curtciftdia and this ia 
almost always the case ; in one or two flowera, hawever, throe ovules bavo Leon 
fonod J Mr. Baker indeed says that the pod of D. Giineifolia may bo S-seedod,— lliis. 
no Calcnttn gpecimGn shows. To D. mofdtnHi Mr. Beatham ascribes aboat 8 
ovalesj" D. m«/acccrtHiH has had, in almost every flower examined, 4 ovules and in 
some pods it has 4 eeoda j one or two ovaries with 5 ovules havo boon met with, 
but never more than 5 have bfien seen. 
The plants named VAii. ? aptera and VAR. ? millet fioides are placed her© merely 
for convenience of reforeuce. They are both reported in fruit only, and as they 
have ahrsost eicactly the leaves of Derrh mataccenifis it seems better for tho present 
to refer to them nnder that Bpeeies, As r^igards VAR, ? aptera indeed this is the 
more essential sinca two gatherings from Perafc (Vfrny 2025) Eunslle)- '3100 \), and • 
ono from F^enang (Curtiti 2735 IJ havo pods intermediate between those of var. ? 
tipterti and those of D. malaceenfiig. The pods of vAft. f nptti'a are, however^ 
obviously those of a Pongamia rather than thoao of u JCKjrrtjn, if Pongamia ha 
really entitled to a separate generic position, which the writer hardly believes. 
The distinguishing character is a quit© artiiicial and, as these very plants show, a 
somewhat inadeqnalo one. 
The exifitenco of VAtt P milleftioides ruisea an even more troublesomo qupation, 
the relationship of Millettia to Den-is. The arrangement adopted iji tho Gcnem 
Plantarm>i, tho Hii^toire des Pliiutes and the Natiirlicher^ Pjlanzenfaiuilien phiccs 
DeiTts find Pongrimin among th^ DalbergiciB and Millettia among tlie Galegeze. This 
then, considering the ^ren^i authority of the authors who have sanctioned it, must 
bo accepted as the moat natural arrangement possible. That a more inconveni. 
ent ono could hardly be devised has, however, been the oxperionco of most field 
botanists and of mostanthors who liiivo had to deal with the species belonging to the 
genera, For these genera uro so cloaely allied that they only differ^ &ad th4*t merely 
