397 
vegetation and on the chresard of the soil. If the chresard is high, it will 
be rapid ; if, as indicated above, it is low, then it will be slow. Imperata 
exaltata grows to the height of a meter or more. This altitude together 
with its more or less dense sod enables it successfully to compete with 
other grasses, but occasionally openings in the “parang” are found in 
which Andropogon aciculatus, which is propably a better forage grass 
than cogon, is present; this forms a carpet, but unlike the cogon it spreads 
by runners above ground instead of by root stalks. It does not grow to 
a great height and many instances were noted where it is being crowded 
out by the cogon. The Andropogon areas are never so extensively de- 
veloped as the cogonales although patches of the former are often found in 
the latter. Imperata arundinacea koenigii, the smaller “cogon” grass, is 
also present in the grass lands, although it, like the Andropogon is not 
so successful as I. exaltata. While many other grasses are present, 
especially in the early stages of the development of the “cogonales,” yet 
none are conspicuous numerically. The large bunch grass, Saccharum 
spontaneum, so characteristic of the lowlands in semi-hydrophytic con- 
ditions, is confined to individual specimens on the rocky flood plains of 
the rivers and to the abandoned rice lands near the mouth of the Alangan 
River. 
Psidium parang.—The invasion into the cogonales of arboreal species, 
or into caingins, brings about the initial stages of temporary formation 
in which tree instead of grass species predominate. After the beginning 
“weed” stage the cogonal, as shown above, may intervene before the tree 
stage is reached, or if the conditions necessary for the cogonales are not 
obtained, the transition may be more or less direct. All the steps in the 
transition from the newly made caingin to the fully developed woody 
type have been observed. If the caingin is surrounded by a forest in 
which there are no exotic species, the seeds of the latter which enter 
will need to be blown by the wind or carried a long distance by birds and 
other animals. The chances are that none or but a few exotics will be 
present. Without exception, clearings observed in the Anisoptera-Strom- 
bosia forest had no exotic tree species. The surrounding forest or the 
closely lying open places furnish the seeds which are to develop into the 
planis which form the invading vegetation. The floristic composition 
will be discussed under a treatment of such clearings in connection with 
the Anisoptera-Strombosia formation. 
As in the case of cuttings, the first “clearings” are made near the 
seacoast, or if inland, they may radiate from a common center (settle- 
ment) in one or more directions. When, finally, the clearing has been 
made and abandoned, as already described, it may pass through the 
cogonal stage; but if the conditions for the development of the latter 
are not present, then the arboreal one has an early beginning. The most 
common tree species found in these clearings and the first one to enter is 
41339—4 
TNs ie aes 
