62 MB.JBENTHA.M OK LOGAlJfIi.CE^-. 



Accordingly in EuloganiecB we may distinguish three groups : the 

 first with a valvate sestivation, corresponding with HeSyotem^ in- 

 cludes the American Spigelia, the American and East Indian Mi- 

 treola^ and the Australian and Indian Mitrasacme\ the second 

 with an imbricate quincuncial sestivation, approaching Scropliula- 

 rinece^ comprises the Australian Logania^ the North American 

 Polypremimi^ the South African Qomphostigma, Nuxia and Chi- 

 lianihuSy and the almost cosmopolitan Buddleia ; the third, with a 

 contorted seativation like that of ApocynecBy is limited to the 

 Asiatic and Australian Oeniostoma, 



-1 



.-I 



^' 



i ^ 



^ t- 



5. Spigelia, Linn. 



This genus, in its habit, inflorescence and broad capsule, closely 

 corresponds with OphiorrMza among Muhiace^. The peculiar 

 dehiscence of the capsule and the articulate style are its readily 

 recognizable characters, and the species are well worked up by Alph. 

 DeCandolle in the ' Prodromus.' Since then, few if any really 

 new ones have appeared, nor have we more than one in the 

 herbaria I have access to. There are, however, several described 

 species to suppress. Some of the perennial herbaceous species 

 will flower the first year, so as in that state to have been described 

 as annuals, and it would seem that the common herbaceous ones 

 have usually a large- and a small-flowered variety. 8. antJiehnia^ 

 a common weed in tropical America, includes 8. ne^^osa and 8. 

 multispicata of Steudel, which are not even marked varieties ; my 

 8. 8chom'biirg1ciana is the same as 8. Sumloldtiana^ and >S^. Mexi- 

 cana seems to be but the larger-flowered form of the same species. 

 My 8. humilis varies in the same manner in the length of the 

 flowers both in Spruce's and in Schomburgk's specimens. 8. gra- 

 cilisy DC, is the same as *S^. spartioides, Cham, et Schl, I ^^ 

 unacquainted with the two Mexican plants of Galeotti, described 

 by Martens as new 8pigelicp, for I cannot find them in Sir W. 

 Hooker's herbarium, whose set is in general nearly complete. 



6. MiTBEOLA, Linn. 



This genus, well characterized by Torrey and Gray and by 

 Alph. DeCandolle, very closely resembles some species of Olden- 

 landia \ but the ovary is free, and the two-horned capsule readily 

 distinguishes it from all others. The styles, separating at the base 

 whilst they stiU adhere under the stigma, recall a very common 

 structure in Apocynea, with which, however, Mitreola has little 

 el«e in common. Pour species are described in the * Prodromus, 







'a 



r 





■-^ 





r 4' 





/ '. 





• . *./-.. 



fl 



' Ov' 



_? 



