244 MI?. G. BENTliAM ON EUPHORBrACEiK. 



it would appear that what Grisebaeh mistook for a single poly- 

 androus flower is in fact a cluster of diandrous ones ; for the 

 stamens are frequently very shortly connected in pairs, and here 

 and there minute scales may be observed at their base within the 



cluster. 



Mueller 



trace) 



of the cluster or cymule. The plant appears to me to be in all 

 respects a true Gymnanthes, and totally unlike any Oleacea, to 

 which order Mueller w r ould refer it. 12. Actinostemon, including 

 Dacfylostemon, is a tropical American, chiefly Brazilian, genus, of 

 which 24 species have been enumerated ; but some are evidently 

 repetitions. The genus has usually polyandrous male flowers, 

 and would technically go into our first group ; but its nearest 

 affinities are evidently with Gyrtinantlies : the principal character 

 distinguishing it, besides the number of stamens, consists in the 

 scarious scales which inclose the very young flower-shoots ; the 

 styles are also united at the base or higher up in a slender 

 column. Mueller distinguishes Dactylostemon from Actinostemon 

 in the transverse embryo ; and certainly in the commonest species, 

 A. (or D.) verticillatus, I have found it very oblique or nearly 

 transverse ; but Baillon insists on the uncertainty of the character, 

 and the seeds of most species have never been observed. At any 

 rate the genera are so absolutely undistinguishable in any other 

 respect, that this character alone w r ould be insufficient to keep 

 them distinct. 13. Adenopeltis, a single Chilian species with the 

 habit and peculiar fruit of Stillingia, from which it only differs 

 in the absence of any male perianth. 14. Colliguaya comprises 

 five extratropical South-American species, resembling Stillingia 

 and Adenopeltis in habit, and, like the latter, without any male 

 perianth ; but the male flowers are collected in clusters as in 

 Gymnantlies pohjandra, and the rhachis of the cluster is adnate to 

 the subtending bract, so that the flowers appear to be inserted on 



the bract. 



Mexican 



distinct in habit, usually lobed leaves and pedicellate male flowers, 

 with some other characters that I have not had the means of 

 investigating for myself, and have been obliged to extract from 

 Baillon's and Mueller's descriptions. 



There remain three American genera, all agreeing with Hippo- 

 manese in, their floral structure, but differing from all other Euphor- 

 biacere in their bracts, and having no direct connexion with each 



eac 



They are : 



