72 



THE BOTANICAL MAGAZINE. 



[Vol. XX Y I. No. sns. 



Combined. 



Here we see that the fruit of the upper part is greater in 

 weight than that of the lower one, as in 1909. On the contrary 

 the number of the vesicles in the former is clearly smaller than 

 that in the latter. 



From these results it seems probable, that the weight of 

 the fruit varies strongly with its position on the branches, 

 while the difference of the vesicles between the upper and the 

 lower branches is not quite clear. We see, then that the number 

 of vesicles is a more fixed characteristic than the weight of the 



fruit. 



Comparing the results of three years (1909-'ll) in A. ha- 

 bitat, the annual changes may be summarised as follows : 



It is to be noticed that the fruiting percentage of the above 

 tree was very good in 1909; poor in 1910, but in 1911 it was 

 pretty good. We see, therefore, that the number of vesicles 

 fluctuates correlatively with the fruiting percentage, and that 

 it is large in a certain year, then less in the following. If we 

 compare the difference of the annual variations with that of 

 the lower and the upper parts in the same year, we know that 

 the former exceeds the latter, because the greatest difference 

 shown by the former is 0.861, while that of the latter is 0.577. 



I shall now give the results in C. habitat. 



