118 



the Journal of Indian botany. 



The peculiar behaviour of the second clusters of No. 3 and 8 is 

 apparently due to cold. The distinction between two clusters is easily 

 determined by the formation of the scale leaves at the end of each 

 cluster. The formation of the first cluster of No. 9 was thus spread 

 out over one whole month, with a break in the middle. 



1917 



1 Continuation of the last year's cluster- 

 In both years the position of the asterisk, under the first, second 



or third letter of the month indicates that the cluster unfolded in the 



beginning, middle or end of the month respectively. 



It would be seen that the number of the clusters formed by the 



ten plants during the two years were as follows :— 



It is clear that the largest and probably the oldest plant formed 

 only one cluster each year. The rest formed one, two or three, and 

 all of these formed more than one in at least one of the two years. 

 From these observations and others given further on in the paper 

 made on much older trees which were growing on land in the garden, 

 it will be seen that in young plants more than one cluster is usually pro- 

 duced each year, while older plants produce only one cluster every 



