228 ^^i^ Philippine Journal of Science ' isis 
that are found in the Archipelago. Collections have been made 
in most of the larger islands, and it has been very rare that any- 
comprehensive collection received at the Bureau of Science has 
failed to present some new form or forms of Medinilla, While 
a few species of the genus are of wide Philippine distribution, 
such as Medinilla myrtiformis Triana, M. rami flora Merr. (per- 
haps both best considered under the genus Anplectrum) , and 
M, a^tronioides Triana, others are very local, and a very large 
number of species are known but from a single locality. Most 
species are found at medium and higher altitudes, and almost 
without exception in the forested regions. Very few species 
are found at low altitudes, and then only in those regions having 
a heavy and continuously distributed rainfall. Most of the 
species are terrestrial, but a number are found in the mossy 
forests as true or facultative epiphytes. 
At the present time, including the forms considered below, 
about 100 species have been described from Philippine material, 
more than are known otherwise in the entire genus over its 
extended range from tropical eastern Africa, the Mascarene 
Islands, India, Malaya, the Marianne Islands, and Polynesia. 
The number of Philippine forms will certainly be considerably 
increased as exploration progresses, for at the present time 
there are several very distinct forms in the collections of the 
Bureau of Science, represented by imperfect material, which 
cannot be referred to any described species, while collections 
recently received present still additional apparently new forms 
which will be studied later. 
With the exception of two or three species, it seems to be 
evident that practically all the Philippine forms are endemic, 
but in spite of the very high percentage of endemism, it is 
apparent that the Archipelago presents a center of distribution 
for the genus. The genus is not particularly highly developed 
in the Malay Peninsula (12 species), or in Java (18 species), 
and so far as the flora of Borneo is concerned (17 speoies) it 
apparently agrees with that of Java and of the Malay Peninsula 
in having comparatively few representatives of the genus. Com- 
paratively speaking, however, the flora of Borneo is very im- 
perfectly known, and the same is true of the islands to the 
south and southeast of the Philippines, so that any conclusions 
drawn from the present known range of Philippine species may 
later have to be radically revised when the neighboring islands 
are botanically more thoroughly explored. It is suspected, how- 
ever, that Borneo will eventually yield a great many additional 
