353 



boki, 16 occur also in the Philippines (but 6—10 of them are 

 more or less common throughout the whole archipelago). Many 

 of them have also been discovered in countries, situated between 

 the Sunda-isl. and the Philippines, viz. Celebes and the smaller 

 islands southward of it (cf. Premna timoriensis, P. pubescens, 

 P. littoralis, V. glabrata, Clemdendron longiflorum).. 



In the case of Clerodendron longiflorum and C. longituba 

 (q. v.), there is also a distinct relation to New-Guinea. 



Further there is a close relation between the Verbenacex of 

 the Philippines and those of Java, Sumatra and Celebes, etc., 

 whilst the relations east- and north-ward arc much less conspicuous. 



It is also a remarkable fact that nearly as many species are 

 common to Borneo and the Malay Peninsula, as to the latter 

 and Sumatra, the distance between the former being 450 — 500 

 KM., that of the -latter hardly 50 KM., these countries being, 

 moreover, joined by a chain of islands. 



If it were not that we had to our disposition only the data 

 of one family we should be inclined to say that there is a 

 direction of plant-dispersal which is more distinct than the 

 direction perpendicular ,to it, viz. the N.-E.^ — S.-W. -direction. 

 It seems even, that the Philippines, which have by far the 

 largest number of Verbenaceaespecies, ate a centre, from which, 

 more or less radially ^), the plants are dispersed towards Malacca, 

 towards Sumatra, towards Java, towards the L. Sunda-isl., via 

 the countries between the last-mentioned and the Philippines and 

 that in this wise, that the more species are common to it and 

 to the Philippines, the more the radius is directed towards the 

 continent of British India. Indeed, we observe from the above- 

 given lists that there are relatively few species common to two 

 countries, situated on the circle, the centre of which is in the 

 Philippines, e.g. Borneo and Celebes. 



As most species of the Verbenaceee possess drupaceous, 

 fleshy fruits of an often bright colour, we could suppose a 

 dispersal by birds (cf. Hallier f.. Med. 's Rijks Herb. Leid. no. 



') It is also possible that the direction is a mere N.-E,— S.-W. one, 

 since Clerodendron longifiorum and C. longituba occur both in N.-Guinea 

 and in, respectively, Timor and Lombok. 



23 



