GIRTH INCREMENT OF HEVEA BRASILIENSIS. 
83 
of new shoots on any given tree, one at the end of April and 
the beginning of May, and the other at the end of June and 
the beginning of July. But all the branches of any given 
tree do not produce new shoots at the same time, and conse- 
quently their appearance is prolonged, so that the two periods 
overlap, and the second is indefinitely extended. On a tree 
which showed new growth in the leader on June 23, the lower 
branches were elongating on August 25. Counting the first 
of these dates as the beginning of the third burst of the year, 
this period of growth was spread over two months. The 
individuality ” of the separate branches has here full plaj^ 
and it is this which has given rise to the exaggerated idea 
as to the number of times these growth bursts occur during 
the year. On any given tree of this species there may be, 
including the first growth after the leafless period, three 
outbursts of new shoots during the year at Peradeniya. On 
the other hand, many trees in the immediate neighbourhood 
may not produce any new foliage after the first. In the case of 
a leader three new lengths may be produced in the year, but on 
the lower branches it would appear that at most only two occur. 
It will be noted that in several of the curves a decrease in 
girth is indicated during the dry months. As the unit adopted 
is only a millimetre, it might be considered that this decrease 
is only an apparent one, and due to errors of measurement. 
But as the measurements were always repeated whenever no 
increase was shown, and as the decrease continues in several 
cases for some weeks, it is believed that the figures indicate 
an actual occmTence. In 1912 this decrease varied from 1 
to 3 millimetres, and was exhibited by twelve trees out of 
the sixteen. In the wetter season of 1913 it was in no case 
greater than 1 millimetre, and was shown by only seven 
trees. In 1914 it occurred in four trees out of ten, and varied 
from 1 to 2 millimetres. There was no scaling off of the 
bark during these periods, the pencil mark which served to 
indicate the place of measurement remaining distinct through- 
out. It would appear, therefore, that a slight shrinkage of 
the stem can occur in the dry season. 
It has been stated that Hevea brasiliensis responds rapidly, 
by alterations in girth, to differences in the amount of water 
