CLASSIFICATION. 



229 



lob 



the lower 



being 



nutlets of considerabl 



slightly th 



longer. 



Tl 



oblique hilum : the nutlets are 



th a fleshy 



e fruit consists of drupaceous or baccate 



ely) corky pericarp and a broad, somewhat 



h 



& 



§ EUGOMPHOSTEMMA 



they are hispid with 



ally glabrous, though in most cases where I 



in 



POGOSIP 



hair 



it is 



All four nutlets develop as a rule 



Stenostoma only one of the four nutlets 



al for one or more to be aborted 



shap 



broadly ovate 



in 



one group (Parvifiora) they are 



becom 

 oblong with generally the 



•rmally developed 



The 



ovary 

 in 



in 



uul 



d 



up } 



d 



d 



(G. eriocarpon) they are globose. The colour of the 



d 



end wider than the lower 

 upwards 



j 



m 



two species (G. lucidum, oblongum) the fru 



utlets has be. n 



ttl 



cue species 



e 



is 



dark red or orange. The 



white, 



i 



in 



m 



are 



ually somewhat unequal 



brv 



ther (G. melmifolmm) it is 



is straight and hardly fills the nutlet, the cotvled 



Ther 



are 



distinguishable in th 



to 



hardly deserve 



with distinctive names and definition 



be recognised as S 



L'rnUS 



thr 



dit 



Although 



or *n«f, whit h. however 



in the t< xt tin 



i 



this has been <i 



rati 



pi- 



led 



convenience than because sectional or even subsectional 

 divisions are : 



for the sake of 



k is claimed for them. 



a 



B 



Pogosiphon, which includes species where the corolla tube is hirsute within 

 this it is usual for one or more of the nutlets to be abortive: 



> 



m 



Stenostoma, where the throat of the corolla is comparatively narrow and the tube 



is nearly straight; in this it is very unusual for more than one nutlet to be d 



e 



loped : 



and 



Eugomphostemma, with an incurved tube as in Pogosiphon, but glabrous within 



four nutlets usually become developed. 



le second (Stenostoma), 



species into which the 



very much longer and larger; in this all the 



The most compact and natural of these divisions is* tl 



Much more natural than these divisions are the " groups " of 



genus separates itself. In some instances (e.g., Parvijfora, Lucida) it might be convenient 

 to deal with the constituent "species" as only somewhat distinct forms or varieties of one 

 very variable plant. In the meantime, however, and until the whole area occupied by the 

 genus shall have been more fuily explored, it seems preferable to give descriptions of the 

 whole of the separable forms as yet reported and (provisionally at least) to recognise 



1 



these as " species.'* In doing this the lines laid down by Mr. Bentham and by Sir Josepl 

 Hooker in their admirable accounts of the genus have been followed as closely as possible 

 But while this has been done, it may not be out of place to indicate briefly the reductions 



the material at present available sugg 



Th 



which a somewhat prolonged study of 



can be best shown in a tabular presentation ; the details are given in each case in the 



text that refers to the group involved. 







X 





/ 





Ann. Hoy. Bot. Garb. Calcuita Vol. III. 



