>h kiwi.-. I BELVISIAI 



111:111 



tin \ have Lin 



there turned inward* ; the anth 



comes a deep !!• -h\ cup oi 



which the narrowi si 



baide. The ovarj i- Inn 



m< 11-. an. I .11-;. : !■ ! 



axile placenta, which :- -•■ attached t.. tl 



tin' hollow centre of tb< 



are oblong, uiili a di pri asion in tin- mi. I. IK- on 



the nucleus being curvi .1 liki a 1 



contact ; tin- style i- 5-angl d, or rath, r 



Minn:,, with fi 



are w rhaps the n 



by ih.- calyx, a- ! 



palp whi.-h it eatable, and a rind bo full ol u nnin, that I 



The Beeda are large amygdaloid bodies, ki.ln.;. 



their contraction the plano-convex cotyl 



an. I plumule ar.- l...tli immers .1 in the - 



In the total absence of all corr< el informati 

 genua, Botanists wi r. unable t.. 

 All that they had been able t • settle was it 

 Palisot De Beauvois Btated (1807) that, in tic 

 Order between Cucurbits and ra-i.iiitlwu.r-: a \ 



[uence of the double-ringed corolla, which 

 1 raflowi re, and the plaited corolla with an ii 

 the flowers of Cucurbits. Desfontaines, on thi conti 

 genus which he calls Asteranthus, without any dou 



1 lopetalous perigynous corolla, its i in the 



oblong two-o !1. .1 antli.r--. single Btyli . 



stem, an. 1 alt.-rnat.- has.-. Him followed Dr, 



and Asteranthus into an < >rder called Belvisiete, wit hoi 



rte position in the Natural System. II- obji 



doubted it- affinity t.. Passionflowi re, an ! c ni] . 



Latterly le seems to have attempted 



Iliilli.ii. 1- jmts it next t.. Symplocacesc ; Mi ism r n< \- 



luri been previously known ol it, thai it 



originated in Di Beaw - scription of them, "S 



Finally, 1 myself, feeling that these could 



panal Alliance, with m doubt. I 



the kin. In. bb of the l'.arl ol Dl ii.; . 



field, from which the 1 ription ai 



the / 



It i- obvious that Napoli ona has nothii 

 been referred. Prom Cucurbits it differs utterly 

 placentation, highly develop. ! corolla, and whole 

 t.i that Order. Passionflowi 1- aeem at first si§ 

 because of the triple-rowed corolla ol V, 1, which m 



those plants ; but there thi 

 ovary, distinct styles, polypetalous corolla, in 

 most essentially at variance with th> e 

 the monopetalous corolla, indefinite epipctalouf 

 and defini 



wholly adherent, with a great epigyn 

 no albumen, t.> say nothing of the 1 1 

 whollj disn garded in a consideration ol this kind. I'] 

 true affinity i-- in the neighbourhood .■!' tl 



•1- : The ovary is in both 

 have a coriaci ras vah . calj \ : 1 th I 

 The placenta of Kan. 1. ha i- almost 

 genus the petals are broken up into nui 

 genus in question. To this maj 1 

 N oleona and of youi ; 

 of slender acicular tubes, which gn th •- 

 containing slender bristles. Finally, the rih ; 



