or retards the formation of roots and shoots in the notches. In the isolated leaf of _Bryophyllum calycinum_ no callus formation takes place and hence no flow of the sap away from the leaf will occur. This will allow one or more of the notch buds of this leaf to grow out and then a flow will be established towards these growing buds. [Illustration: FIG. 18 FIG. 19 FIG. 20] In the third specimen, Fig. 18, the presence of two leaves suppresses or, as a rule, retards the growth of a shoot on the stem and possibly also the flow from one leaf may block to some extent the flow from the opposite leaf if the piece of stem is very short. This puts the leaves in a condition not as good as that in leaf Fig. 20, but better than in leaf Fig. 19.[153] [153] With larger leaves the experiment may also succeed in moist air. In the normal plant the buds in the notches of the leaf remain dormant since the flow of the “stimulating” substances takes place towards the tips of the stem and root, and because these substances are retained there in excess. This is probably the real basis of the mysterious dominance of the “whole” over its “parts” or of the anlagen of the tip of the stem over those farther below. When a piece of the stem of _Bryophyllum_ is cut off and its leaves are removed, the two apical buds will grow out first. This “dominance” finds its explanation probably in the anatomical structure and the mechanism of sap flow which tend to bring the “stimulating” substances first to the anlagen in the tip. In _Laminaria_ Setchell has been able to show directly that regeneration always starts from that tissue which conducts nutritive material. When we cut out a piece of a stem of _Bryophyllum_, and remove all the leaves, new shoots will be formed from the two apical buds of the stem, and roots will arise from the most basal nodes; provided that the stem is suspended in air saturated with water vapour. The growth in such a stem deprived of all leaves is slow. If we remove all the leaves on such a piece of stem except the two at the apical end, the stem will form only roots, but these will develop much more rapidly than on a stem without leaves. If we remove all the leaves except the two at the basal end, the stem will only form shoots (at the apical end) but these will develop much more rapidly than in a leafless stem. Hence the leaves accelerate the growth of roots towards the basal end and inhibit it towards the apical end; and they favour the growth of shoots towards the apical end and inhibit it in the nodes located nearer the base.