80 LKGUJtKOSS. 



racemis axillaribus ^-^-fasciculatis |)edicellis 1 ciii. lougis floribus 2-o- 

 fasciculatis, bracteis oljsoletis, petalis ad marginem purpnrascentibus. Calys 

 breve cylindricus, apice subtruncatus, 5-dentatus, dentibus posterioribus 

 subobsoletis, 3-anterioribus conspicuis, inedio i)lus conspicun. "Wsilluiu 

 obovoideum apico emarginatum ; alfe leviter longiores medio cniu caiinis 

 tenuiter conniTentes ; carinre a]3ice liberre, basi distinctfe. ytamen Texillo 

 oppositura liberum, cfetia connata. Stjlus fijiformis f)ostice recurTUs, antice 

 deoi-sum ciliolatus ; stigma terminale, punctiforme ; oTarium sessile, OTulis 5. 

 Fructus oblongi subcomplanati ± obliqui 4 cm. longi 3 cm. lati 'Jo mm. 

 crassi 1-spermi exalati lignosi extus dense tuberculati, proiie suturam 

 sublseves et obscure costati, in matmn rarius ad suturam inferiorem 

 deliisceutes, apice rotundati ad summum breTC apiculati, basi obtusi Tel 

 brcTiter stij)itiformes angustati medio sensira et obscure tenuissime Tix Tel 

 liand constiicti, pedicellis 1 cm. longis Talidis. Semina obreniformia 3 cm. 

 longa 2icm.lata; testa nitida nigiicans, ad apicem medio et ad centrura pio- 

 funde impressa ; cotyledonibus araplis late oblongis 3 cm. latis 22 mm. longis 

 basi profunde eordatis, sinibus 14mra. longis; radicula obsoleta. 



H.U3. SankakuTii, leg. K. K\g.u, 1902. 



The present ijlant is one of those leguminous plants whicli aflbrd a 

 poisonous matter nsed for fisliing by the aborigines in Foimosa. It was 

 first labelled by Prof. J. M.^TSUiruR.v as Derris cMnemis Bexth. Soon after, 

 the professor regarded the plaut difierent from Bentham's species and 

 named it D. taiioaniana. Tlie same plant was studied by Dr. K. Nag.u 

 who regarded it to lie referable to LoncJiocarpus or raore probalily to 

 Cmtbulandia, but not to Derris. He then expressed this opinion iu a 

 number* of the Journal of the Tokyo Chemical Society. In my opinion, 

 the plant certainly does uot beloug to Derris on account of its •vvingless 

 fruits ; nor should it be refen-ed to Lonchocarpus, for the plaut has a sessile 

 ovary, but has neither elongated i^od nor broaded snture. It is also difierent 

 from CoidiiiJandia in having ueither toralose nor rounded pods. A genus to 

 which the present plant should probably be refen-ed is, iu my opinion, 

 Pongamia, as is seen from the structnre of the frnits. The descriptions of 

 • Joumnl o£ the Toky5 Chemicnl Society XXIH — 7, pp. 744-777 (.Tapnnese). 



