NATUfvAL ORDERS. 151 



Magnoliaceoi and CedrelecB, wliicli arc common to 

 America and India, liavc not heen fonnd on tlic continent 

 of Africa, nor on any of the adjoining Islands. 



GuUifercB and llesperidece exist, tliougli sparingly, on 

 other parts of the coast. 



A few plants really belonging to MeJiaccce have been 

 found on other parts of western eqninoctial Africa, and a 

 species of Leea (or Aquilicia, for these are only different 

 names for the same genus) which was formerly referred to 

 this order, occurs in the herbariinn from Consfo. 



M. de Jussieu, who has lately had occasion to treat of 

 the affinity of xVquilicia,^ docs not venture to fix its ])lace in 

 the system. Its resemblance to Vinifera) in the singular 

 structure of seeds, in the valvular aestivation of the corolla, 

 in the division of its leaves, the presence of stipules, and 

 even in inflorescence, appears to me to determine, if not its 

 absolute union, at least its near affinity to that order. Of 

 VinifercE, Vitis is at present the only certain genus ; for 

 Cissus and Ampeh}ms having, as Richard has already 

 observed, exactly the same structure of ovarium, namely, 

 two cells with two erect collateral ovula in each, should 

 surely be referred to it ; nor is there any part of the 

 character or description of Botria of Lourciro, which pre- 

 vents its being also included in the same genus. 



Lasianthera of M. de Beauvois,^ referred by its author to 

 Apocineac, but which iM. de Jussieu has lately sug- [kjc 

 gested may belong to Vinifera?,^ is too imperfectly known to 

 admit of its place being determined. 



forming; a third section, from the remarkable evolution of its embryo and con- 

 sequent want of albumen. Id the ajstivation of calyx and corolla it ac^rccs 

 with Jacquiiiia, which together with Theophrasfa^ (or Clavija of the Flora 

 Peruviana), forms the fourth section ; characterised by the squama?, more or 

 less distinct, of the faux, of the corolla, and by generally ripening more than 

 one seed. The fifth, includes only Btcobotrt/s of Forstcr (the M<€sa of Forskal) 

 which, having ovarium inferum and five barren filaments alternating with the 

 segments of the corolla, bears the same relation to the other genera of this 

 order, that Samolus does to rrimulacca.\ On the near aflinity, and slight 

 differences in fructification, between this family and Myrsincas, 1 have formerly 

 made a few remarks in the Prodr. Plor. Nov. lIoU. 1, \). 53:^ 

 ' Mem. (Ill Mas. iVliist. Nat. 3, ;j. 4-37 ct 441. - Flore iVOicare, !,_/;. S5. 



^ Lor. cit, 



