IN CEYLON. 
455 
The general facts may be summarized as follows :— 
(1) A high temperature in the tropics compared to Europe 
is not necessarily accompanied by an increase in the percent¬ 
age number of deciduous arborescent species {of. Colombo 
and England). 
(2) A decrease in temperature accompanied by a marked 
climatic periodicity is associated with an increase in the 
percentage number of deciduous species, if the tree flora of 
Great Britain is compared with that of Ceylon. 
(3) A decrease in temperature in parts of Ceylon where 
the climatic periodicity is nearly the same does not lead to 
an increase in the number of deciduous species ( cf . Hakgala 
and Peradeniya). 
(4) An increase in temperature in other parts of the same 
tropical island is not necessarily accompanied by a pro¬ 
portionate increase in the number of deciduous species 
(of. Colombo, Mannar, and Peradeniya). This appears to 
contradict (3). 
(5) An increase or decrease in temperature in the same 
district (Peradeniya) is not concomitant with an increase 
or decrease in the number of deciduous species. At Pera¬ 
deniya the temperature gradually rises from February to 
May, and the number of deciduous species first increases 
and then steadily declines. In May the temperature is at 
the maximum and is on an average 3*6° F. higher than in 
February, but the number of species which are either leaf¬ 
less or undergoing defoliation in May is five times less than 
in February. If the temperature and number of deciduous 
species throughout each month of the year be considered, it 
will be seen that from January to March there is a rise in 
temperature and an increase in deciduous species, a further 
rise in temperature from March to May with a decrease in 
the deciduous species, a fall in temperature and a decrease 
in the number of deciduous species from June to January. 
(See plate XXYI.) 
These facts seem to suggest that in parts of Ceylon, and 
probably in the tropics generally, the effect of a high 
