372 BOTANICAL GAZETTE [NOVEMBER 
Balsam, at such elevations, rarely matures its seed, reproduction being de- 
pendent on seed blown up from below and on the layering of the trees them- 
selves. Branches so layered are often found growing as independent trees, 
the connecting branch having decayed. The rooting always proceeds from 
rmant buds. Prostrate balsam occurs at an altitude of 5500 feet on the 
Presidential Range, where it reproduces almost entirely by layering. At an 
elevation of 4900 feet cones are borne sparingly, but even here reproduction 
by layering exists. 
II. Layering as observed on Isle Royale, Lake Superior 
Upon Isle Royale the layering habit manifests itself as follows. 
In the forest one frequently comes upon small groups of young 
balsams, often of about half a dozen individuals of various sizes. 
These upon superficial inspection would easily pass for a cluster 
of seedlings, but if the group be carefully dug up, all the young 
trees will be found to be connected with each other a little below 
the surface of the ground. The way in which the layering comes 
about was found to be as follows. The lower branches of the balsam 
droop more or less, and the middle portion of such a branch fre- 
quently becomes covered with moss, litter, and humus. If the 
point of origin of the branch is very close to the ground, the connec- 
tion soon becomes entirely concealed; this seems to be the case 
more often than otherwise. The covered portion now produces 
roots abundantly, and the tip becomes erect, loses its dorsiventral 
character, takes on radial symmetry, and is to all appearances @ 
perfect miniature tree. 
_ Layering may take place at any stage in the life of the tree. 
Sometimes the layered branch may be only a few years younget 
than the parent and not very perceptibly smaller, showing that 
it must have developed from one of the very earliest branches. 
Mere seedlings were sometimes seen with layered branches about 
as large as the parent. The daughter trees often produce a second 
generation, and it is in this way that the groups of apparently 
independent saplings come into existence. On the other hand, 
cases were found where a mature tree was layering through branches 
that had their points of origin a number of decimeters above the 
ground. Several of the lower branches of a mature tree may layer, 
producing a circle of young trees around the parent, and numerous 
