MR. G. BENTHAM ON EUPHOUBJACEJS. 239 



triandrous male flowers, differ but little from each other ; Agirta 

 and Lassia, each containing one Madagascar species, also trian- 

 drous, are unknown to me. Leptolotrys, the common North- 

 American species, is almost always diandrous. Leucandra, origi- 

 nally proposed by Klotzsch as a genus for a tetrandrous Brazilian 

 species with ascending but not twining stems, has been extended 

 to comprise a few other Brazilian species with the normal twining 

 habit and 4 to 8 stamens in the male flowers. Bia, witli 10 to 20 

 stamens in the males, has four South- American species, including 

 Leptorhachis of Klotzsch, which in the l Prodromus ' is associated 

 with Gtenomeria as a distinct genus, but restored by Mueller to 

 Tragia in the i Flora Brasiliensis.' Ctenomeria, a South-African 

 twiner, also polyandrous, but very dissimilar to Leptorhachis, 

 forms a distinct section of Tragia. ZucJcertia, from Mexico, 

 another polyandrous species, is unknown to me ; Baillon. retains 

 it as a distinct genus. Lastly, Seidelia is so very different in 

 habit and character from Tragia that I cannot understand the 

 principle upon which Mueller includes it in that genus. Baillon, 

 as already mentioned, has correctly understood its affinity to 

 Mercurialis and Adenosine. 11. Cnesmone, a single tropical 

 Asiatic tall climber, is near to but sufficiently distinct from Tragia, 

 and is equally ambiguous between Plukenetiese and Acalypheae. 

 12. Daiechampia, about GO species, chiefly American, but ex- 

 tending also into tropical Africa and Asia. The genus is per- 

 fectly well defined, but has the habit and characteristic style of 

 Plukenetieae, and is in some respects nearly allied to Tragia, though 

 often widely distanced from it, and formerly placed in Euphorbiese 

 as having the flowers involucrate. To this character Mueller 

 always attaches great importance, although the so-called invo- 

 lucre is often very different in homology. In this case it con- 

 sists of two of the lower bracts of the short, often capitate inflo- 

 rescence being much enlarged and often coloured ; but even this 

 character is not constant; for in the South-American D. micrantha, 

 for which Klotzsch proposed the generic name Rhopalostyles, the 

 inflorescence is looser and the lower bracts not at all or scarcely 

 enlarged. 



Subtribe 8. Hippomane^e. This is one of the most natural 

 subtribes of Crotonese, generally admitted but difficult to define, 

 and variously limited by different botanists. It was well marked 

 out by Adrien de Jussieu as his fifth section, to which Bartling, 



t 2 



