houstoun's central ajierican leguminos^. 229 



Bentbam in Kjoeb. Vidensk. Meddel. 1853, 13. Bentham bad not 

 seen Jacquin's type, nor bave we; tbere is, however, a fragmentary 

 specimen in Herb. Banks from bim (from Cartbageua), iu which 

 the leaflets are narrower than in Seemann's Panama specimen 

 (no. 202), which Bentbam identifies with the species, and with 

 which Houstouu's plant agrees. In Jacquin's figure of the fruit 

 the costfe are more marked than in Houstoun's specimens. The 

 plant in Herb. Sloane 292, 22, agrees with Miller's specimen, but 

 a second sheet in Herb. Banks froin Houstouu may perhaps belong 

 to a different species, but the material is insufficient. Houstoun in 

 his MSS. gives a full- description of the fruit, which it may be worth 

 while to transcribe : — 



"Arbor est foliis, floribus, magnitudine, et toto habitu, Colutea; 

 vnlgaii similis, singular! vero siliquarum forma multum ab ea 

 discrepans. Hujus etenim siliqu;e pollices binos cum dimidio sunt 

 longffi, octo circiter lineas latie, vix ultra duas crassse ; et ita com- 

 presssie ut diameter major ea sit qua3 valvarum media jungit, minor 

 qute earundem margines. Intus autem nou unica est cavitas, sed 

 margines valvularum counectuntur septo transverso, duplici con- 

 stante membrana in qua latent semina plura oblonga, reniformia. 

 Utrinque vero est cavitas, dicto septo et tumente valva formata, 

 soloq. jere repleta." 



Plukenet's "Colutea Verne Crucis vesicaria" (Aim. Ill, t. 165, 

 fig. 8), referred here by Miller, is not this plant. 



CoRONiLLA scANDENS. This plant is probably Chcetocalyx jnibescens 

 DC. (Prod. ii. 810), but DeCaudolle's description is insufficient to 

 determine this. C. pubesctns is not given by Grisebach in his Flora 

 of Brit. W. Indies: the plant was originally described from St. Do- 

 mingo, and has not, so far as we know, been taken up by authors. 

 Bentham named as C. pubescens a specimen in Herb. Kew from 

 Fendler, Venezuela, No. 291, and there is placed with it another 

 specimen from Galipan (Murit::, No. 14), with which the Miller 

 plant agrees very well. There are, however, on Fendler's plant 

 small scattered prickles similar to those of C. vincentinus, which we 

 do not find on Houstoun's specimen. Miller's plant in Herb. Mus. 

 Brit., which was placed by Bentham iu ChcEtocaly.r, but without 

 specific name, does not seem to correspond with any other described 

 species ; should it prove distinct, it might well take Miller's name 

 as Chatocali/x scandens. Miller gives an excellent diagnosis and 

 description in English of the plant, which he grew iu Chelsea 

 Gardens from seeds sent him by Houstoun from Cartbageua, and 

 which he considered identical with Coronilla scandens pentapltylla ,oi 

 Plumier. On this latter Linnaeus based his Coronilla scandens, also 

 a plant of uncertain affinity ; but Plumier's figure {PI. Atiwr. ed. 

 IJunn. t. cvii. fig. 8) does not correspond with Miller's plant, though 

 it may be one of the Brazilian species of the genus. 



Hedysarum purpureum = Desviodium tortiiosion DC. (Meibomia 

 tortuusa O.K.). Miller's ticket is missing, and Houstoun's specimen 

 from Vera Cruz is not, as usual, written up with Miller's name ; 

 but the synonymy and description leave no doubt as to the identity 

 of the plant. 



