53 



FLORA TELLUR. 



distinct by petals, style, anther. I had called it 

 Malacarya in 1817. 



950. Xypuostylis R. diff. Canna, cor. basi 

 tubulosa, segm. 3 int. ineq. undul. filam. ineq. 

 bipart. anthera adnata, stylo ensiformis, stig. 

 lin. involute — Types 1. X. lutea Raf. Canna flac- 

 cida Rose, and Amer. Authors, fol. ovatobl. fl. 

 luteis — 2. X. angustif. R. Canna do W. auct, 

 fol. lanceol. fl. rubris, cor. segm. int. rectis 

 emarg. Both from Florida to Louisiana. The 

 real G. Canna (root of Canna-bis) or rather 

 Katubala of Adanson has style clavate, stigma 

 obtuse and other differences, 



D51. CRATODIA Raf. 1815. This order 

 of mine distinguished from Orchides by a calix 

 and fruit commonly multilocular, was the Nat. 

 fam.ily Scitaminea of Linneus ; but this name 

 has been restricted to the Gr. Musa, Heliconia 

 and akin with berries <Slc by Ventenat, The 

 Gini^embres of Adanson or Cratophytorum of 

 Necker included also the Bromehdes properly 

 separated by Jussieu. But many Nat. fami- 

 lies are included in the Cratodes, besides Sci- 

 tamines, such as 



Drymiriiizes of Ventenat, Types Amomum, 

 Maranta, Alpinia, ^c Curcumides a subfamily 

 with several stamens. 



Stratides Raf. 1815, types Stratiotes, Hy- 

 drocharis &c, and perhaps Aspidistra with Ma- 

 croiryne. 



The Valisnerides and Pistides with unilo- 

 cular fruit, may with the Cratodes that have 

 such a fruit, be formed into a link with Orchi- 

 des, as a Sub Order of either, or perhaps a pe- 

 culiar order Synarmes Raf. 1815. To revise 

 all the akin Genera would be a task, at present 

 I shall merely indicate two old Genera of Adan- 



