VOCIIYSIACE.^. 99 



oj^posite sirajjle leaves, accompanied by caducous stipules, and flowers 

 disposed iji more or less raiaifiod and compound termimal racemes. 



Liffhtia,^ which connects the other Vochijsiaccce to Trigonia, is nearly 

 allied to the latter, but distinguished from it by its receptacle which 

 is more concave, and its perigynous petals reduced to three, the 

 fertile stamens being only four in number, didynamous, and the 

 ovary cells only containing two ovules. Two arborescent species ^ 

 have been described, one from Guiana, the other from the Amazon. 



This small family, distinguished in 1820 by A. S. IIilaiee,^ 

 has been considered by some as allied to the Combretaceœ and 

 Onagrarieœ^ because of the perigynous character of most of its 

 genera ; by others to the Geraniaceœ, on account of the analogy often 

 presented by its spurred flowers to those of the Nasturtiums and 

 Pelargoniums ; Lindley, considering them as inseparable from the 

 Poli/galaceœ, with which he even connected the Triyoniem. By the 

 intervention of the latter, the Vochijsiaceœ should perhaps be placed 

 in the same family Avith the Polygalacece, representing the perigynous 

 series of them. They would be distinguished by their mode of in- 

 sertion, although the concavity of the receptacle, so decided in Light ia^ 

 disappears almost comi^letely in most of the Trigonias ,- and still 

 more so by the irregularity of the androceum, normally reduced to 

 a single fertile piece in all the Vochysieœ and Erismece. An affinity 

 has also been indicated between the Vochysiacece and the ^apindacece. 

 It is principally manifest by the intervention of the Chailletiece with 

 irregular flowers like Tapura. When, however, we know the close 

 relationship of the latter with the Euphorhiuccœ proper, especially 

 with Pedilanfhus, whose irregular flower much resembles that, of the 

 Vochysiacece^ we can understand that Trigonia has often been con- 

 nected with the family Euphorhiacew^ 



The characters of the three series which we distinguish in this 

 family are the following : — ■ 



1 ScHOMiî. in i/H««-(7, XX. 757.— B. H. Gen. Spec. i. 123 (1824).— Exdl. Gai. 1177, Ord. 



977, n. 7. 260.— Vcchi/nceii: IjiSDi.. Iiitrod. eà. 2, ti7 \ Viij. 



■ Walp. Ann. i. 190. Kingrl. 379, Ord. 134. 



' Injrém. 3lm. vi. 253 [Vuchlsié-n). — Viichy- ''The Clusiaccœ and the MarcgraiUir [Ti-rn- 



sieœ DC. Prodr. iii. 25, Ord. 69.— E. Mey. in strœminceœ) have been compared with the 



Nov. Act. Nat. Cur. xi. 812. — Bartl. Ord. Nat. Vocht/siuceiB, hut we see scarcely any affinity 



320.— H. Bn. in Piu/er Fran. Nat. 350, Fam. between them. 

 155. — Vuc/n/siacctc Mart, et Zl'OC. Nov. Gen. et 



