THE FLORA OF MANILA. 171 
small, for species it averag-es approximately 40 per cent, and 
the indications are that as botanical exploration progresses this 
percentage will be increased rather than diminished. 
As noted above the average percentage of endemism for the 
entire Archipelago is about 40. In many families it runs to 
90 per cent, or even higher, but in others it is much lower than 
the average. Wherever primeval conditions exist, that is, wher- 
ever the original forests have been undisturbed by man, the 
percentage of endemism is always high but wherever the coun- 
try is thickly settled, the destruction of the original forests and 
the large introduced element, cultivated plants, weeds, etc., cor- 
respondingly decrease the percentage of endemism, so that we 
find in Manila a percentage of endemism of only 8.7, or including 
cultivated endemic species, only 12 per cent. This great differ- 
ence in endemism emphasizes the radical differences between the 
vegetation of the settled areas and that of the forested regions, 
and is one strong reason for considering a high percentage of 
the species found in the settled areas as introduced. 
In support of the theory that very many of the species now 
common and widely distributed in the settled areas in the Phil- 
ippines have been introduced into the Archipelago since the 
advent of man, and in many cases have been spread in the 
Archipelago by his agency, the present vegetation of many of 
the smaller islands can be cited. On very many of such islands 
which are not inhabited, and only visited at irregular intervals 
by fishermen, or by collectors of turtles, turtle's eggs, or bird's 
eggs, it is exceedingly rare that one finds a single species that 
has presumably been introduced by man, although the conditions 
for their growth and reproduction are frequently present. I 
have personally visited a number of small and isolated islands 
in the Sulu Sea, and while many of these support a fairly rich 
flora, the species, without exception, are those that have been 
distributed by natural causes, that is, by water, winds, or mi- 
gratory birds. The characteristic species of the open country 
on the larger islands are practically wanting, although as noted 
above, favorable conditions for their growth on these small 
islands are sometimes present. However, as small islands in- 
variably support a flora poor in species as compared with larger 
land masses, the evidence from this source alone cannot be of 
great value. 
The case of the Island of Polillo is of interest in this connection, 
for it ranks about sixteenth in size among the islands of the 
