ANONACEjE. 247 



BoUiiiia, and Bocagea. R. Brown liad in 1820 establislied the genus 

 Artabofri/s, for those Ohl World species of Uoaria and Unona in which 

 the principal axis of the inflorescence is like a flattened fasciated 

 hook ; and the genus which he had made known six years earlier, 

 under the name oi Eupomatia, though long held of doubtful aflBnities, 

 and unfortunately pointed out by A. L. de Jussieu, as the type of 

 a new order near Osyridece, was already accepted by several botanists 

 as nearly allied to the Anonacece. In 1832, A. de Candolle was led, 

 in a special work on this order, to break up the genus Xyhpia, which 

 is now reconstituted, and to propose two new generic types — 

 Milima of Leschenault and the monopetalous genus Hexalohus. So 

 there existed at that time sixteen of the genera retained by us in the 

 order Anonacece. The twelve others are of quite recent creation. 

 Between 1832 and 1866, English botanists made known the genera 

 8ager(Ba, Cyathocalyx, PhcBantliUS, SpIi(Brofhalamus, Dlsepalum, Cym- 

 bopetalum, Cleisfochlamys, Enantia, Atndreyia. Endlicher had named 

 Popoiviain 1838; A. Richard, Oxandra in 1850. Miquel is the 

 latest author who has studied the Anonacea of tropical Asia in situ ; 

 he established the genus Tetrapetalum in 1866, thus raising the 

 number of genera received by us to twenty-eight. But of these 

 it is probable that some wall be suppressed when transition terms 

 shall be better known, that will allow us to admit them as sec- 

 tions into several of the older genera. We have fortunately not 

 been compelled to establish any fresh generic type in this order. 



On seeking out the characters constant in all these groups we 

 find that there is no Anonad that is thoroughly herbaceous ; 

 that all have alternate exstipulate leaves, and in the seed contain 

 fleshy ruminated albumen. 



Other important characters are so common in this order that 

 their absence has only been made out in one single genus thus dis- 

 tinguished from the rest. These are as follows : — 



I. The form of the floral receptacle and the resulting insertion 

 of the androceum. — Only in one type, E/ipomatia, has the flower a 

 totally concave receptacle, with its stamens all inserted above the 

 gynseceum. 



II. The presence of petals and sepals. — In Eitpomafia alone are the 

 sexual organs surrounded by a simple bract, which falls off'by its base, 

 and plays the protecting part of a perianth, this being really absent. 



III. The independence of the carpels. — In Monodom alone are 



