246 NATURAL HISTORY OF PLANTS. 



base so as to free the fertile and sterile stamens that push it off.' 

 We may, then, define the genus Eupomafia as Anonacea with naked 

 flowers, in which the perianth is replaced by a single modified leaf, 

 and the carpels are inserted on a concave receptacle. It is in this 

 order analogous to Tiochodendron amongst the Magnoliacea. 



AnoiKB form Family XL VI. of Ad an son's great work ; this, as 

 we have seen, includes not only those Anonacece that were then 

 known, but also MagnoUacea, MeniapermacecB, several Dilleniacece and 

 B(inuncuJacea>, Ochna, and Fagara. Of this group the genera wliich 

 really belong to Anonncea are four in number — viz., Anona, Xi/Iopia 

 {Xglopicron), IJvaria {Nanim), and A-simuia. Adanson was the first 

 to recognise the analogies between Xylopia and Aiiona ; and his genus 

 Narum includes both Uvaria proper and also the Asiatic Unonas of 

 the group Cananya. The Anonce of A. L. de Jussieu only include 

 the five genera Anona, Unona, Uuaria, Cananga, and Xglopia. Most 

 of the other genera united to these by Adanson, he reserved for his 

 order Magnoiuicece. L. C. Richard gave the collection the name of 

 Aiionacece, and this order was only really established in the work pub- 

 lished by Dunal in 1817, so wonderful considering its date. To the 

 genera above enumerated are there added the following : — Kadsi/ra, 

 which belongs to Schizandrea ; Monodora, whose type is the Anona 

 MijriHlica of G.ertner ; Porcdia, which Ruiz & Pavon had made 

 known in 1794, and Giiatter'ui of the same authors, corresponding to 

 Aublet's Cananga. Dcsmos and Mdodorum, proposed as distinct 

 genera by Loureiro in 1790, are by us incorporated with the great 

 genus Unona. A. P. de Candolle, in 18:24, fully adopted the 

 arrangement o{ Anonacece proposed by Dunal. Soon after, Blu.me 

 compU'tely revised most of the Old World genera, assigned more 

 exact limits to the existing genera C^/?o«aand Uvaria, and established, 

 either as distinct generic types or as sections of other larger genera, 

 the groups O.rgitiitra, jMUrrp/iora, and Orophaut, whose autonomy we 

 maintain. About the same period, A. de Saint-Hilaire was doing 

 the same work for the American Anunacco', and successively created 

 the genera Anavagorea, Duguetia {Abcrcmoa of AuBLET, 1775), 



' Tlie flowor* only IiimI. n dnv. afttr wliiih tlio into ii sort of rinj: inserted near the «lgo of 

 whole net of Hiiiinenn, Hterilo iin<i ffrtiie.eoiiio off the receijtaculur cup, 

 in a 8in(;le ciicular piece, their hiutoHhcing united 



