. 
194 _ - The Philippine Journal of Science 1917 
a heterogeneous mixture in which many plants occur in situations 
very different from those in which they are usually found. The 
occurrence of patches of Miscanthus, a typical high-mountain 
grass, near the sea level is a conspicuous example. Equally strik- 
ing is the growth of Ipomoea mixed with Saccharum at consider- 
able distances from the shore. The scattered patches of Mis- 
canthus in an unusual situation can hardly be regarded as con- 
stituting an association. The same might be said of Ipomoea 
that is growing between the clumps of Saccharum. The scat- 
tered plants of Ipomoea and Canavalia that are found on the 
beach can hardly be considered as a strand formation. Cer- 
tainly there is little similarity between the growth composed 
of these species on Volcano Island and the normal strand forma- 
tion as found on the sea beach. : 
The great preponderance of grass on the island seems to 
justify the classification of the vegetation as an invasion by a 
grass formation. The principal association of grass is certainly 
Saccharum spontaneum. In some places on the steep slopes 
Themeda gigantea occurs in stands that are pure enough to 
justify its being classed as a separate association. On the main- 
land both Saccharum and Themeda constitute very definite 
associations. 
The tree flora is so scattered and is composed of such a hete- . 
rogeneous mixture of different species that it is impossible to 
recognize any clear divisions, and the whole can best be regarded 
as an early stage of the invasion of grassland by a second-growth 
forest formation. The prominent species are all small, and all 
are characteristic of the early stages of the invasion of grass- 
land or other open areas by trees in many parts of the Philip- 
pines. Second-growth forests should give place to the tall dense 
forests characteristic of this region. In this process there must 
be a number of stages or successions. Our knowledge of these 
successions is very fragmentary at present, so that the different 
stages cannot be described. However, on Taal there seems to be 
no indication of a second stage, as all the species of any promi- 
nence are characteristic of the very first stages of the invasion 
of grassland or other open areas by second-growth forest. 
Gates ** divides the invading trees and shrubs into two forma- 
tions. One of these he calls the parang, or shrub-small-tree, 
formation and the other, the low-altitude tree formation. In 
each of these he recognizes a single association. The first is the 
” Gates, F. C., Philip. Journ. Sci. 9 (1914) Bot. 391-434, pl. $-10. 
