IN CEYLON. 
457 
period, the rains preceding and following this miniature 
drought apparently being such as to obviate the necessity 
for a more general leafless condition. 
At Hakgala, though there is a drought during February 
and March very similar in intensity to that at Peradeniya, 
yet the vegetation is by no means strikingly deciduous, and 
if it were not for the few leafless trees of Meliosma Arnot- 
tiana, the introduced fruit trees, and species of Cedrela, 
Michelia, &c., it might be described as typically evergreen. 
Nevertheless there is a response to the period of drought in 
so far that a large number of the deciduous trees pass 
through their leafless phase at that time. 
Similarly it can be shown that in parts of the Uva and 
Northern Provinces the periods of drought, though occurring 
at widely different times of the year compared with Pera¬ 
deniya and Hakgala, are coincident with the fall of leaf from 
a large number of trees. But in all these cases it should not 
be forgotten that the minimum monthly rainfall is coinci- 
dent with the minimum monthly humidity of the air—a point 
which suggests that either factor may be equally potent in 
leading to the production of a temporary deciduous character 
in the vegetation. 
In comparing various districts in Ceylon which are widely 
different from Peradeniya in their annual rainfall, say 
Hakgala with 91*7 inches, Badulla with 76*93, and Mannar 
with 38*59 inches, it will be noticed that the annual quantity 
of rainfall is not arithmetically related to the number of 
deciduous species in each district. For instance, Hakgala 
with a heavier rainfall shows less deciduous species than 
Peradeniya, but Badulla and Mannar, though they have less 
rainfall than Peradeniya, do not necessarily show a larger 
number of deciduous species, and it is therefore necessary to 
consider the united influence of humidity of the air and 
rainfall. 
