188 
BULLETIN OF THE NATURAL HISTORY SOCIETY. 
bank here is rapidly washing away (no doubt because of a pro- 
gressive sinking of the coast), 
and this had evidently deprived 
the tree of support when very 
young. It has accordingly sunk 
down, at the same time bending 
geotropically upward, until it 
now rests against the bottom of 
the bank, while the main root 
ascends vertically upward for 
some six feet (the tree itself 
being about that height), and 
then bends over horizontally into- 
the soil at the top of the bank, the present approximate outline of 
which is shown by the curved line. Whether or not the tree has 
taken root also in the bank at the base, my notes unfortunately 
omit to state. This case shows two interesting facts ; first, that 
the washing away of the banks must be at times very rapid, 
(though apparently intermittently), and second, it shows that 
woody plants must retain their powers of geotropic bending much 
later than is commonly supposed, and than the nature of the tis- 
sues would lead us to expect. 
The third figure represents a form of spruce very common on 
the elevated open barren plateaus just east of the valley connect- 
ing the Xepisiguit and Upsalquitch. All such trees, which grow 
only to about ten feet in height, show a lower very dense part 
extending up a foot or two from the ground, 
a living pyramidal rather close top, and a 
branchless bare trunk between. This form 
is due, I think, to the following causes. 
Growing as these trees do in the crevices of 
the rocks of windy bare plateaus, they mu?t 
at times, especially on the bright, warm, 
windy spring days, be subjected to an intense- 
transpiration when the water supply is very 
limited, or, in the early spring, still frozen, 
and hence unavailable. At this time the 
lower branches are protected from excessive 
transpiration by the snow covering, and in. 
