1916] LOEB—BRYOPHYVLLUM 299 
the upper leaf. To settle this question control experiments were 
made in which both leaves were equally near the water (fig. 6); 
this case both leaves formed shoots, while when they were os 
higher up in the trough neither formed roots and shoots. Both 
leaves also formed roots and shoots when 
put horizontally into the water. The influ- 
ence of moisture upon the growth of roots 
and shoots.in the notches was very strikin, 
in all these experiments. Thus contact 
with the wall of the trough, where a con- 
densation of water kept some of the notches 
moist, would favor the growth of roots 
and shoots in these notches. 
It has been stated by previous writers 
that in a completely isolated leaf of Bryo- 
phyllum the notches 
grow out because 
the stalk of the leaf 
does not form roots. This statement is not 
strictly correct, as has been stated already 
in the first article. An example was given 
in fig. 16, p. 261, where a leaf formed roots 
and shoots in one notch although roots and 
a shoot had been formed on the stem, owing 
to the fact that the piece of the axillary bud | 
was still in contact with the stalk of the leaf. 
We must assume that in this case the suction 
from the new growth at the stalk of the leaf 
was less than if a piece of stem had been 
there, and this lessened suction permitted 
one notch of the leaf to grow into a shoot. — 
It occurred to the writer that this assump- 
tion might be put to a test. Leaves were 
Prepared being connected by a piece of the cortex of the stem, 
without the wood (fig. 11), and the behavior of such leaves was 
compared with the behavior of leaves which were attached to a 
complete piece of stem (fig. 12). Both sets of leaves were suspended 
Fre. x3 
Fic. 14 
