322 MR. H. H. W. PEARSON ON THE 
of a moor-pan, the effect of which in preventing the penetration 
of the roots of trees is too well known to need description; 
there is, however, at present no direct evidence of the existence 
of such a formation beneath the humus-patanas. 
Lastly, it may be pointed out that bacteria do not flourish in 
acid humus (33), and that the consequent poverty of such a soil 
in combined nitrogen would co-operate with other factors against 
the establishment of a luxuriant vegetation. 
Summa ry. 
An examination of such evidence as exists with regard to the 
origin of the patanas of Uva and their western extensions up 
the slopes of the central ridge leads to the following conclusions, 
On the Uva slopes below 4500 feet (the lower limit of the 
Rhododendron) the peculiarities of the climate have co-operated 
with the periodically recurrent grass-fires to transform an open 
forest of low xerophytic trees with an undergrowth of grass 
( e. a Savannah-forest such as is still found on the eastern 
boundary of the plateau) into barren grassy plains. These 
plains being almost completely denuded of soil must be regarded 
as being of the nature of a permanent savannah, the natural 
re-afforestation of the greater part of which is impossible under 
the present climatic conditions. Above 4500 feet, wide tongues 
of patana extend in a westerly direction up to, and in some 
cases over, the summit of the central ridge. There can be no 
doubt that these extensions are due to the encroachment of the 
Uva grass-fires into the montane wet-zone forest. Upon the 
cleared area a herbacecus vegetation has established itself, 
the remains of which form an accumulation of sour humus which 
is almost uniformly present on the surface above 4500 feet, 
The properties of sour humus are such that forest-trees can 
with diffieulty re-establish themselves upon it. It therefore 
follows that, apart from the effects of the present annual fires, 
the sharp boundary, once established by fire, would so gradually 
become irregular by tbe advance of forest.growth that only 
careful observations extended over a long period would be able 
to detect any change. Hence has arisen the idea that the present 
limit of the forest is a stationary one. 
