395 
height if they are given an opportunity to develop and they are always 
found as taller trees in the bamboo forests or in more xerophytie physio- 
graphic situations within that forest. 
The interpretation of the genetic degeneration of the Bambusa-Parkia 
formation given above is based almost entirely on comparative observation. 
Different stages from the lowest to the highest type may be seen lying 
adjacent to each other in a single habitat. These places usually adjoin 
the more reduced stages, viz, the cultivated clearings and cogonal. In 
two instances the progress of the vegetation on actual cuttings was 
observed at intervals throughout a period of fourteen months. Both, 
abandoned before completion, were made in the Streblus-A phananthe 
type with the evident intention of clearing the land for cultivation. 
Many of the larger trees were left and in some places in the same clearing 
only slight inroads were made on the standing vegetation. An observation 
of these stages at later intervals, showed an incomplete development of 
all the forms of the parang just described. 
It will be noted that no mention has been made of the vegetation 
occupying total clearings. It has been seen how the forest gradually 
has been reduced to the parang stage. It is only one step further to 
destroy it entirely. In the case of the parangs discussed, the forest is 
not cleared and prepared for cultivation; while in that of clearings the 
whole of the vegetation may be removed, the land tilled for a number of 
years and then abandoned, or what is more often the case, patches of it 
may be denuded, or large trees left standing. It will at once be seen 
that the resultant nature of the vegetation which invades these clearings 
when they are once more left to nature depends on their condition of 
reduction. Let it be assumed that the reduction is complete. Grass 
lands may then predominate in such clearings and a discussion of the 
causes of their development will first be given. 
Cogonales.—Clearings in the Philippine forests or on their borders, 
which are covered by grasses, are known as “cogonales” (PI. V), but there 
are no places on the Lamao forest reserve where the latter appear as the 
climax vegetation. Without exception, they are only stages in the reéstab- 
lishment of the normal vegetation removed by man. ‘The first periods in 
the development of vegetation in caingins, in the best expression of the 
Bambusa-Parkia forest, have not come under my observation, but it 
is not thought that they differ materially from those in the parang 
after it has been well established or indeed in clearings made in the 
Anisoptera-Strombosia forests to be discussed below. Usually, such 
clearings are made on the borders of others and consequently, when 
abandoned, many seeds from these places find early lodgment. However, 
if the clearing is made during the dry season the migrated seeds do not 
develop until the wet one. This gives an opportunity for seeds to migrate 
before many germinate and therefore the first season shows vigorous 
weeds such as Blumea balsamifera, Elephantopus mollis, Emilia flammea, 
