In the preceding cases the rudiments of the leaf-buds and of the roots were probably present in most cases, so that gravity only awakens them into activity. In other forms, as, for instance, in heterocentron, it is possible to show that gravity may even determine the production of new buds. If pieces of the end of a branch, including the growing point, are suspended vertically, some with the apical end upward, others with the basal end upward (Fig. 34, _A_), the former produce roots only around the base, but in the latter roots appear frequently, not only at the base, but even extending along the stem. They appear not only at the nodes, where pre-formed rudiments may be present, but also in the internodes, where there are no rudiments of roots. Stems of heterocentron placed in a horizontal position produce a circle of roots around the base, and later, in several cases, roots from the under surface of the stem, both from the nodes and the internodes; but these roots are smaller than those at the base. Those around the base are often longer on the lower side than on the upper side. Vöchting has also studied the regeneration of pieces of roots of the poplar and of the elm suspended horizontally in a moist chamber. A callus develops from the cambium region of the basal end, and from this a thick bunch of adventitious sprouts grows out. A weak callus may develop on the _apical end_ also, from which a few roots develop. In other cases adventitious shoots are produced also from the _apical callus_, especially from the upper edge of the callus. The results are variable, but show that at times leaf-shoots may develop from the apical end of the root. It is also singular to find that, while pieces of the root produce new leaf-shoots very readily, yet they often fail to produce new roots, or produce only a few that arise from the apical callus or from the sides near that region. It is difficult to show that gravity has any influence on the result.