TKOMVTIoN 





XI 





dioecious sub-group being again subdivided into sections which an* founded 

 number of tlie stamens and the situation of tho receptacle r & . Ilf tl|| 



sections into which I have thus thrown he Indo-Mala a specios, I hav 



on tl 



M 



M\ 



ni 



l« 



ectional designations wor.ls previously in use as sectional or subgauric in. - ]• 



fiwt section, as already stated, I have invented a new ..an.-, wl.ick imli,*te. what I 

 believe to be its posi [on in the evolution of tho genus; and for the seventh I ,ve al 



ented a new name, indicating its newnoss in point of evolution. The urru. i.ant 



If 



as follow 



GROUP L— Pseudo-hermaphrodite: male 1 .wers with 1 stamen and ft rudimentary pistil 



Pseudo-hermaphrodite flowers sad gall nW<n in 



one set of receptacle*: fertile female flowers 



another set p i 



GROUP II.— Unisexual or asexual ; male flowers without rudimentary pistfls 



Section L— Male, gall, and fertile f. inal flowers on the same 



(rofnorphf. 



receptaele 



Section II.— Flowers unisexual or neuter: male and all flowers on 



one set of receptacles, fertile female and 1 tit r 

 flowers in anoth- r set 



Section III.— Flowers unisexual: male and gall flowers in one set 



of receptacles, fertile female flowers only in another 



Urogtigtnn. 



'vnnecio. 



■ 



I 



set 





A.— Flower monandrous 



a, Receptacles chiefly axillary .... ftidixm 

 , Receptach J mo tly in fascicles from 



stem ad branch* i ( 



B.— Flowers di-, rarely triandrous 







a, R. ceptacles mostly axillary 1 



£, Receptacles mostly in fascicl s from stem 



U*yc< 



and branch' i JS mmphe. 



These seven sections are no. all equally natui I. The most natural of them ore Uro$fgma 

 and Syncecia. The coincidence in Urostigma of such apj irently unconnected characters as 



the monoecious condition of the axillary ] tired receptacles and the epiphytal 1 d t is ven 



remarkable. In no other section is the tendency t be epiphytal at II strongly n rked 



in Urostigma it is universal. Many species in other t> tl BS are scandent and support 



themselves on trees and rocks by throwing out rootlets from (heir si m« and branch- 



But these rootlets are furnished with fibrillar and collecting hairs like the rocAa that 

 penetrate the soil, and are very different in appearance from the str ng sul livisions 

 of the main axis by which the epiphyte embraces, and ultimately sir ogles, the ti ) t » 



which it attaches it elf. The name Urostigma was originally devised by Gaspanini. 





