\fR. Q. BETfTHAM OK AXOXACEJE. 67 



Fig. 3. More advanced stage— the cotyledons nearly equal. 



Figs. 4, 5. 8. biflorus (with nat. size of larger cotyledon). Cotyledons already 

 very unequal. 



Fig. 6. The same. 



Fig. 7. The same, a : scar of a detached leaf («) ; b, smaller cotyledon. 



Fig. 8. S_. polyanthus -. a, inverted leaf; 6, young undeveloped scape; c, flower- 

 ing scapes ; d, enlarged foliaceous and persistent cotyledon. 



/ 



Notes on Anonacecn. By George Bextkam, Esq., F.L.S. 



[Read May 3rd, I860.] 

 The history, affinities, and generic distribution of tins very natu- 

 ral family have been so fully discussed by Hooker and Thomson 

 in their ' Flora lndica,' that little remains to be observed on the 

 Asiatic species, from the materials we at present possess; but 

 there is still some degree of confusion with regard to the American 

 ones. "When Martius, in an early part of the great ' Flora ' pub- 

 lished under his editorship, described the Brazilian Anonacece, the 

 principles more recently adopted for their classification had not 

 yet been proposed ; and although the principal American genera, 

 Guatteria, Duguetia, Anona, Xylopia, Bollinia, Bocagea, and 

 Anaocagorea, remain pretty nearly -within the limits there ascribed 

 to them, the generic characters of some of them, as well as their 

 classification in tribes, require modification, and the Uvarias of the 

 ' Flora Brasiliensis' do not at all correspond to the Asiatic genus of 

 that name. In our ' Genera Plantarum,' Dr. Hooker and myself 

 purpose following, with some slight modifications, the tribes pro- 

 posed in the 'Flora lndica,' which, although somewhat artificial, are 

 the best which have suggested themselves. We would, however, 

 suppress the Anonece, uniting them with Xylopiece ; for the charac- 

 ter derived from the carpels consolidated in the fruit is variable 

 both in Bollinia and Duguetia. We would also somewhat extend 

 the character of Mitrepliorece, so as to include Oxymitm, Phtfanthus, 

 and Monodora, in which the inner petals, although not unguicu- 

 late, are erect and connivent or coherent, whilst the outer dis- 

 similar ones are spreading, as in other 3Iitrepliorece. The Guat- 

 teriecB, which cannot include the genus Guatteria as originally 

 constituted for the American species, must take the name of 

 Unonece ; and Popoida, admitted to be intermediate between these 

 Unonece and the Mitrepliorece, appears to us to be better placed in 

 the former than in the latter tribe. 



The true American Guatterias, numerous in species, all belong 

 to the tribe of Uvariece, as characterized by the imbricate aestivation 

 of the petals. The inner petals are very decidedly so in all the 



