ix, c, 3 Robinson: Distribution of Philippine Mosses 201 
the case. It is perfectly true that hundreds of species are 
known each from a single island, on which there is at present 
every reason to believe that they are endemic ; but this is equally 
true of localities on a single island. It appears to be well estab- 
lished that the general course of migration of plants and human 
beings into the Philippines has alike been from the southwest 
and the south; on the other hand, there are good reasons in 
discussing the floral provinces of the Philippines for starting 
nearly at the north. 
Much of northern Luzon is highly mountainous, the so-called 
Mountain Province being separated from the China Sea on the 
west by a narrow coastal plain and terminated on the east by 
the valley of the Cagayan River. Very many species are known 
or known in the Philippines only in this region; many others 
occurring there are also found on the tops of the mountains 
farther south, such as Mariveles, Banajao, or even Apo. Still 
others have a more continuous distribution, following down the 
ranges of the eastern or the western coast or both, their range 
terminating at Mariveles, or in the hills of Rizal, eastern Laguna, 
or Tayabas, or extending still farther to the south. It is open 
to discussion whether such species are to be regarded as more 
properly belonging to the Mountain Province but with more 
southern extension, or whether the opposite is the case. 
Apart from the Mountain Province, there are two rather 
distinct but by no means absolutely separate plant-provinces, in 
close correlation with the distribution of rain throughout the 
year. The eastern coast of Luzon shares with the greater part 
of the Visayan or central islands and with the southern islands 
a very equable rainfall, the dry season being comparatively 
short; the western side of Luzon has a prolonged dry season. 
The ranges of mountains forming the divide lie much nearer 
to the eastern coast than to the western, but are broken in 
various places, so that the division is far from complete. The 
general result is that there are two fairly definite areas of 
distribution in accordance with these facts, although the dif- 
ference is not so great, nor the lines so sharply drawn, as in 
various other countries, as for instance between the northern 
and southern slopes of certain of the islands of the West Indies. 
It has seemed to be of some general value to determine whether 
the facts believed to have been established for the flowering 
plants hold equally true for the lower groups, and for several 
reasons the mosses have been selected for the comparison. 
Thanks to the courtesy of Doctor V. F. Brotherus, who has had 
a wealth of material from all parts of the world for comparison, 
