Mat 1891.] BOTANICAL SOCIETY OF EDINBURGH. 



113 



Having ascertained these results for the trees taken 

 together, it remains to ascertain whether they hoki true in 

 them individually, and on inquiry it turns out that the trees 

 are divisible into two classes, both agreeing in a general way 

 in the relation of girth-increase to top-shoot growth for the 

 first four months of the growing period, but differing 

 materially in the last two months. In the one class (Table 

 v., A.) the girth-increase rallied to perhaps its greatest 

 vigour in the last period, when top-shoot growth became feeble ; 

 while in the other (Table V., B.) there was no such rally. 



Table V. 



Considering that the girth-increase in April was a mere 

 trifle, the relation between girtji-increase and top-shoot 

 growth in the six trees was roughly as follows. In May, 

 when the girth-increase was very vigorous, top-shoot growth 

 was only beginning ; in June and July, when girth-increase 



I 



TKANS. EOT. SOC. VOL. XIX. 



