OF THE UILLENI4CE^. *" 7H 



in Malacca, W. Scortechinii in Perak, and W. KunsUeri in 

 Perak. All the last three may be supposed to have split off 

 from W. pulchella, which might also give rise to the subgenus 

 Capellia, of which there are two species. W. suffruticosa from 

 Malacca to Sumatra and Borneo, and W. obloncja m the Malay 

 Peninsula and Sumatra. 



Finally, we come to Dillenia. In the subgenus Eudillenia we 

 have D. indica common from Ceylon through the Western 

 Peninsula, the Himalaya, Assam, the Eastern Peninsula, Java, 

 Sumatra, Borneo, &c., wliile most of the other species are local, 

 D. retusa in Ceylon, D. hracteata in Mysore, D. ovala in Penang 

 and Perak, and through Cochin-China to Borneo. These are 

 quite simply explained by mutation ; let D. indica enter from 

 the south and split off the others. D. ovaia was perhaps the 

 first, or the best adapted, for it has spread comparatively 

 widely. 



In the subgenus Colbertia we have one or two groups of 

 species, and I am not quite sure to which one or two of the 

 species belong. It matters little, however, for the purpose of 

 this paper. The widely spreading species of the subgenus are 

 D. aurea and D. pentagyna, the former running from Nepal to 

 Bhotan in the Himalaya and through the Eastern Peninsula to 

 the Andamans, Java, and Cocliin-China, the latter from Oude 

 to Assam and Pegu and Cochin-China, and down the Western 

 Peninsula, but not in Ceylon. Now the species nearlj'^ related 

 to the former appear to be D. ptdcherrima in Pegu, D. scabrella 

 in Assam, and D. 'parvifora in Tenasserim Pegu and tlie Anda- 

 mans, while species nearly related to D. pentagyna are D. grandi- 

 folia in Penang and Malacca, D. florihmida in Martaban, 

 D. reticulata in Perak, and D. Hookeri, D. elata, and D. Blan- 

 charii in Cochin-China. There remains only D. eloiigata in 

 Sumatra, of whose relationships I am quite uncertain. 



The Dilleniacese are very marked in Northern Australia, but 

 it will be at once noticed that the genera which occur there, 

 mainly Pachynema and Hibbertia, can be easily derived froni 

 Tetracera, which runs down to New Caledonia, where Hibbertia 

 occurs also. 



Having now tracked the Dilleniacese through the Indo- 

 Malayan region , we may apply the same theory of mutation to 



7(8)07 i^) 



