302 BOTANICAL GAZETTE [ocroBER 
leaves. Other leaves of the same plant are suspended in moist 
air (fig. 17). Figs. 16 and 17 were drawn after 22 days. The 
leaves suspended in air (fig. 17) had formed vigorous roots (R) 
and a shoot (S) at the cortex. The leaves themselves had formed, 
as usual in this case, numerous roots and some tiny shoots. The 
notches of the leaves suspended in water (fig. 16) had formed 
very powerful shoots and roots and this inhibited the growth of 
roots and shoots in the cortex. Needless to say, these experiments 
were all carried out on many specimens, as were all the experiments 
reported in this and the previous paper. The experiments show 
that a vigorous growth in the notches of a leaf can act as a center 
of “suction”? which may prevent the flow of sap to the cortex and 
thus prevent the growth there, if the suction by the stem, or in 
this case the cortex, is not too strong. 
If the whole piece of the stem is left with its two leaves instead 
of the cortex alone, the experiment may also succeed, but it is not 
so reliable since the suction of the stem is greaterin thiscase. Ifthe 
experiment is made with one leaf and a complete piece of a stem and 
the leaf is put partly in water, the leaf cannot suppress the forma- 
tion of the shoot on the opposite side of the leaf, since the suction 
effect of the stem cannot be overcome by the leaf. But when we 
have only a piece of cortex instead of a piece of complete stem (with 
wood), the “‘suction”’ power of the cortex alone is less than that of a 
complete piece of stem, and hence can be suppressed if a vigorous 
stream of sap to the growing notches of the leaves is started, as is 
the case in fig. 16. 
It is hardly necessary to state once more that the term ‘“‘suction”’ 
effect of a growing notch or bud is used only to illustrate the direc- 
tion and relative velocity of the flow. 
ROCKEFELLER INSTITUTE FOR MEDICAL RESEARCH 
Ew York City 
