454 BOTANICAL GAZETTE [JUNE 
even more slender in the sparse stand upon the granite surfaces, 
but in these open situations there were in addition to the short 
branches on the upper part of the trunk longer ones near the 
surface of the rock (figs. 1 and 2), forming a compact mat, none 
of the twigs more than half a meter high. At least one-half of 
the total foliage of the trees was usually upon these prostrate 
branches, and it would seem from the apparent vigor of the leaves 
that an even larger proportion of the work of food synthesis was to 
Fic. 3.—A dead stump of black spruce with a circle of living offspring from its 
layered branches. 
be referred to this lower stratum. The rooting of the trees in 
these rocky habitats was, as a rule, very shallow, and hence the 
massing of their branches reduced the exposure to winds and the 
consequent danger of uprooting. The habit was necessarily con- 
fined to open stands. 
The mat of lichens and mosses which antedated the tree con- 
tinued to thrive under and among the prostrate branches and the 
resulting soil soon buried portions of the lower members of the mass. 
On Pinus Banksiana this was apparently without results other than 
