FLORA. OF THE PHILIPPINE ISLANDS. 289 



had to contend with in the determination of their plants, was the 

 absence of authentic specimens, consequently they had to fall back 

 upon books— Miquel's ' Flora India? Batavae ' and Hooker's ' Flora 

 of British India ' being two of the principal ones ; and in several 

 of the Orders their determinations must be considered rather as 

 approximate than real : for instance, in the genus Eugenia 33 

 species are identified with Indian ones, but in 23 species from 

 Vidal's collection scarcely one tenth are identical. It is much 

 to be hoped that a set of the types of this work may be deposited 

 in some European Herbarium, thus enabling the errors to be 

 eliminated ; for until this is done it will be impossible to obtain 

 a correct idea of the flora. 



In the following paper a census of genera and species are 

 given under each Order, with a column showing the endemic ele- 

 ment ; then an attempt is made to trace the origin of the flora 

 by a comparison with surrounding countries ; this is followed by 

 an outline of the distribution of some of the more characteristic 

 species within the group, and the physical causes which have in- 

 fluenced this distribution ; and, lastly, a list of some corrections 

 made in working up the aforementioned collection, with the 

 descriptions of a new genus and a few new species. 



In the following Census the left-hand column shows the number 

 of indigenous genera and species under each Order, and the right- 

 hand column shows the endemic t4ement. Yillar's census in- 

 cludes the cultivated and introduced plants ; consequently it does 

 not give a fair idea of the flora. My own is chiefly based on that 

 of Villar, with the introduced species excluded and some recently 

 detected species added ; a few which have crept in by error are 

 also excluded. The endemic element is considerable, but the 

 following numbers must be taken as approximate, and as repre- 

 senting the present state of our knowledge ; in Orders requiring 

 revision, the numbers are under-estimated for obvious reasons. 



