the natural Group of Plants called Pomacea. 



their seeds, as opposed to the suspended seeds of most true 

 Rosacea (Xestler's Potentillea). But whatever may be the \ able 

 of this distinction in other instances, it must in the present 

 family be considered of generic importance only: for Dry a*. 

 JValdttcinia and Gcum, with all the habit and other characters of 

 Rosacea', have seeds with the same direction as Pomacea; and 

 certain Cratagi with angular leaves exhibit a passage from one 

 to the other. Nor can the inferior fruit of Pombcete distinguish 

 them from Rosacea with more certainty than the direction of 

 their seeds, as is manifest from the structure of certain genera I 

 shall presently have occasion to propose. It is true that Pyrtff, 

 Mespilta, and some others*, have fruit absolutely interior, or co- 

 hering with the calyx and each other l>\ their whole surface; 

 but in Cotoncaster this cohesion is very partial, and in true Ph<>- 

 tinia scarcely exists in any degree. There is however one cir- 

 cumstance which is universal in Pomacea, and 1 believe does 

 not exist in Rosacea ; namely, that the ovula of the former are 

 collateral, and of the latter, when more than one, vertical, or 

 placed one above the other. This character may therefore be 

 employed to distinguish Pomacea as a section from Rosacea, but 

 can scarcely be sufficient to separate it as an order; especially 

 as the same disposition of ovula, when reduced to a single pair, 

 exists in Spiraa. 



ROSACE ARUM sectio prima Juss. 



(Pomaces Richard Anal, du Fr. ed. Angl. 2.3.) 



I line Myrtaceis baccatis, median tibus Chnrnomele et Cydonia 

 pariter polyspermis, affinis ; inde Rosaceis ceteris per Crata - 

 gos semine peltato. 



O J. 



♦ Cll AH A CTER 



