MR. G. BENTHAM ON EUPHOBBIACE^. 223 



majority are American, extending even beyond the tropics both 

 in the northern and southern hemisphere. There are also a few 

 species both in Africa and in the Indo- Australian region, but mostly 

 tropical. All are connected by distinct characters readily ob- 

 served. The inflorescence is a terminal spike or raceme, the male 

 flowers usually clustered, the females singly inserted along the 

 rhachis, or sometimes a female in the centre of the male cluster 

 or cymule ; in the males the calyx is valvate or slightly imbri- 

 cate, and the petals are rarely wanting ; the stamens indefinite 

 and central, few or numerous ; the anthers adnate, inwardly pen- 

 dulous in the bud owing to the inflection of the filament, and 

 becoming erect as the flower expands — a peculiarity which forms 

 the most essential character of the subtribe. The female flow r ers 

 are more generally apetalous, the styles variously divided, the 

 fruit the normal Euphorbeaceous capsule separating into 2-valved 

 cocci. The arrangement of the numerous species is in a very 

 unsatisfactory state. Mueller established 10 sections exceed- 

 ingly unequal and all very artificial, and yet not always easy to 

 distinguish, each one, even those of only two species, compri- 

 sing the most dissimilar forms. Baillon has in various papers 

 proposed about 22 sections, and some of them apparently quite 

 natural ; but he has given no comparative key to them, and many 

 species do not appear to fit well into any of them ; so that they are 

 practically of little use. Klotzsch separated above 20 genera 

 upon characters usually based on the examination of single 

 species, and deservedly rejected by subsequent botanists. The 

 most apparent characters, derived from inflorescence, indumentum, 

 number of stamens, ramification of the styles, etc., are often very 

 vague ; a satisfactory revision of the divisions can therefore only 

 be effected by great patience and the sacrifice of more time than 

 I can bestow r upon it. I have therefore, in the ' Genera,' merely 

 passed Mueller's sections and subsections in review, comparing 

 them as far as I was able with Baillon's sections and Klotzsch's 

 genera. Of the three smaller genera closely allied to but retained 

 as distinct from Croton, Julocroton, about 20 tropical or subtropical 

 American species, is little more than a section of Croton with a 

 compact spike and more irregular calyx ; Crotonopsis is a slender 

 North-American annual, with a normally 2- or 3-celled ovary, 

 but the capsule constantly small, 1-seeded, and indehiscent ; and 

 Eremocarpus is a single North -American herb, with the pistil 

 reduced from the first to a single uniovulate carpel. 



82 



