of the NiUuml Order of Myrti. 259 



hand, Rumphius tells us the Mimufops h.^^ often as'many as three 

 or four perfect feeds. It is probable, therefore, that the germen has 

 eight cells and eight feeds, moft of which are generally abortive ; 

 another inftance of the nccclTity of ftudying that part in all its pro- 

 grefllve flates. 



The fpecies of Imhrlcar'ta are : 



I. Imbricaria crenuliita, foliis obovato-cuneiformibus apicem 

 verfus crenulatis, petalis calycibufque denticulatis. 

 Jungia imbricata. Gartn. he. cit. 



1. I. ciliata, foliis triquetro-linearibus calycibufque ciliatis, ger- 

 mine pentagono. 



To thefe might be added the tenella of Gaertner, which not hav- 

 ing fufficiently examined, I for the prefent omit. All are natives of 

 New Holland, or New South Wales. 



2. B^CKEA Linn.-~Juff. Gen. 321. 

 OcTANDRiA Monogynia, near Fuchfia and Ximenia, 



Char. Gen. Petala 5. Calyx 5-fidus. Capjula tri- vel quadri- 

 locularis, polyfperma, calyce tedla. 



Juffieu firfl: formed any tolerably jufl: conjeftures concerning the 

 natural family of this genus, to which the defcriptions of Linnseus 

 and Oibeclc by no means lead ; nor, indeed, could it have eafily 

 been referred to the myrtle tribe, without the increafed knowledge 

 of that order which we have derived from the plants of New 

 Holland. There is no doubt, however, that Bachea belongs to the 

 Myrti, and not to the Onagr^e of Juffieu, having the clofeft 



L 1 3 affinity 



