240 



GEOGEAPHICAL DISTEIBUTIOjN". 



cannot for phytogeographical purposes be separated from that species 



ii 



occurs in the same form in Tenasserim 



d G 



endemic, cannot for phytogeographical 



javanieum, 



G 



hid 



i 



? 



purposes be separated from G. oblong 



that extends in various local manifestations to the Malay Pen 



rugosum, 

 though 

 a plant 



and Isthmus, to 



the Andama 

 these Java 



d to the Philip 



rn 



The 



ubjoined table 



g 1 



the distribut 



of 



species 



Table X. — Table of distribution for Java (including Madura) 



Name op Species. 



Endemic. 



* 





i 





< J . phlomoides 



• • • 



1 



(G. flavescens) 



G. parviflorum (1 var.) ... 



1 



. 



• ♦ # 



G. rugosum ... 



• • • 



• • • 



G\ javanicum 



■ 



• • • 



• •• 



1 



Totals of artificial species 



3 



Percentages of ditto 



• • • 



68-6°/ 



Totals from probable natural species 



* V • 



Percentages from ditto 





♦ • • 



• • • 



1 



1 



• •• 



««• 



i 



4 



^4 



757, 



To Tenasserim. 



• • • 



• • 



i 

 4 



I 

 4 



n 



a 



33-3°/ 



i 



2 



16-6 



From the degree of isolation indicated by the endemic percentage in this case, we 



must look upon Java and Madura as constitutin 



j 



. w - u iJUU U ava auu mauura as constituting a separate phytogeographic district, 

 southern India has yielded two species (G. Heyneanum and G. eriocarpon) 



both 



are morphologically very distinct and both are endemic, so that, like Eastern China, 



Southern India must be looked 



upon 



constituting a distinct phytogeographical district 



