BOTANY OF THE CEYLON PATANAS, 309 
formed in districts covered by a luxuriant montane forest; 
except in so far as the decomposition of the hard rock may 
proceed more quiekly under the moister conditions which the 
presence of forest-growth implies. Examination of specimens of 
soil, subsoil, and bed-rock from various parts of the Uva pataua- 
district have yielded no suggestion of the presence of a quartzite- 
formation $, Therefore, although the outcropping of the band 
of quartzite may be, to some extent at least, responsible for the 
occurrence of patana where it was observed by Abbay, this solu- 
tion of the diffieulty will not hold generally, and is certainly not 
a true one for the great patana-area of Uva. 
(iii.) The Grass-fire Theory. 
Reference has already been made to the grass-fires which 
appear periodieally on the patanas, and their importance will 
now be considered. The patana-grasses are very coarse and 
wiry, and in their adult condition are unpalatable to cattle, 
numerous herds of which graze in the district. It has therefore 
been the graziers' custom, probably from time immemorial—and, 
in spite of government regulations against it, is so still—to set 
fire to the grass at least annually, before the bursting of the 
N.E. monsoon, in order to provide a young fresh growth during 
the monsoon rains, upon which the cattle will feed. Early in 
October these fires can be seen blazing in all directions ; they 
often burn continuously for several days, temporarily reducing 
the country to a blackened waste which extends up to the 
very edge of the forest, where shrivelled leaves and charred 
trunks bear testimony against the maintenance of a permanent 
forest-boundary. And, although the effect of a single fire is 
undoubtedly small, it is very evident that the eumulative effect 
of such fires during a succession of years must be to materially 
extend the boundaries of the patanas at the expense of the 
forest. 
That similar fires in other parts of the world cause the replace- 
ment of forest-vegetation by a herbaceous and low shrubby 
formation is so well-known that it need not be insisted upon 
here. Grisebach (15) has described it in India, Junghuhn (16) 
* T am indebted to my friend Mr. H. Stanley Jevons, F.G.S., for kindly 
examining the geological specimens referred to. 
