Another experiment on the same animal gives also a somewhat similar result. If a worm is cut in two obliquely (Fig. 21, _B_) and the regeneration of the posterior piece is followed, it is found that the new material appears at first evenly along the entire cut-surface. It then begins to grow faster on one side (Fig. 21, _b_), and a head appears in this region with its axis at right angles to the cut-edge. As the head grows larger the growth is more rapid _on one side_, and as a result the head is slowly turned forward (Fig. 21, _b_). This more rapid growth on one side brings the new head finally into its typical position with respect to the rest of the piece. The end result of these changes is to produce a new worm having a typical form. If the oblique cut is made behind the old pharynx, as in Fig. 22, _A_, the new pharynx that appears in the new material along the cut-edge lies obliquely at first, indicating that the new median line is very early laid down in the new part, and connects the middle line of the old part with the middle of the new head. As the region behind the new head grows larger and broader the pharynx comes to lie more and more in an antero-posterior direction, and finally, when the new part is as broad as the old,[62] the pharynx lies in the middle line of a symmetrical worm. These results show that the new growth may even take place more rapidly on one side of the structural median line than on the other, and on that side that must become longer in order to produce the symmetrical form of the worm. Here also we find that a formative influence of some sort is at work that regulates the different regions of growth in such a way that a typical structure is produced. The more rapid growth on one side is, however, in this case clearly connected with the relatively smaller development of the organs on that side, and perhaps this same principle may explain all other cases. If so the phenomenon appears much less mysterious than it does when the growth is referred to an unknown regulative factor. _DOUBLE STRUCTURES_ A structure that is single in the normal animal may become double after regeneration, and in some cases the special conditions that lead to the doubling have been determined. Trembley showed that if the head of hydra is split lengthwise into two parts, each part may complete itself and a two-headed form is produced. If the posterior end of a hydra is split, an animal with two feet is made. It is true that the two-headed forms may subsequently separate after several weeks into two individuals, and even the form with two feet may lose one of them by constriction, as Marshall and King have shown. Driesch has produced a tubularian hydroid with two heads by splitting the stem partially into two pieces. Each head is perfect in all respects, and although each has fewer tentacles than