MR. G. BENTHAJVf ON EUPHORBIACKiE. 207 



these have the inflorescence of many Crotoneae, and a very few 

 Crotoneae have the typical inflorescence of Phyllantheae. The 

 calyx is valvate in two or perhaps three genera of Phyllantheae ; 

 the petals are fully developed in a few r species ; and a few genera 

 have the indefinite central stamens characteristic of so many 

 Crotoneae. There are also a few genera with the uniovulate cells 

 of Crotoneae, which have nevertheless the outer or single series of 

 stamens opposite the sepals, as in Phyllantheae. In order not to 

 invalidate the characters of the two great tribes, I have collected 

 these last exceptional genera into a small intermediate tribe, the 

 Galeariece. The main character also, the difference between the 

 uniovulate and biovulate cells, is in a few cases not quite certain. 

 There are some biovulate genera where one only of the two ovules 

 is fertilized, and as the ovary becomes enlarged after flowering 

 (the state in which it is usually examined) the uuimpregnated 

 ovule remains exceedingly small and has often been overlooked. 

 The three tribes must therefore be regarded as to a certain degree 

 artificial, and may hereafter be proved to be more so, and made 

 to give way to a better arrangement ; but in the present state of 

 our knowledge they appear to form the most useful, as well as 

 the most generally received primary division of this part of the 

 order. The subdivisions of the two great tribes may be more 

 natural; but those at least of the Phyllantheae, which I propose 

 first to consider, are as yet too indefinite to be classed as distinct 



eubtribes. 



Taking, first, the petaliferous genera with a normal Phyllanthous 

 inflorescence and uniseriate stamens, we have a group of six or 

 seven, remarkable for their large fleshy cotyledons, with the 

 albumen reduced almost to a membrane, or entirely deficient. 

 All have also the rudimentary pistil well developed in the male 

 flowers, a character brought prominently forward by Adrien de 

 Jussieu, and recently by Mueller, and very constant in many 

 instances, though not always to be relied on. Two of these 

 genera, Bridelia with about 25 species, and Cleistanthus with 

 about 22, all from the Old World, have been removed by 

 Mueller to a distant aubtribe, on account of their valvate 

 calyx ; whilst Baillon, making little accouut of that character, 

 unites them as sections with the American and West-African 

 genus Amanoa. It cannot be denied that there is close affinity 

 between the three ; but the characters which separate them 

 are constant and definite, and appear fully to warrant their main- 



k2 



