[Illustration: FIG. 42.--_Planaria lugubris._ _A._ Two heads produced after operation similar to that in Fig. 24. Each head about half size. _B._ Worm split in half through level of pharynx. New half-worms larger than half of normal worm.] the head that regenerates from an undivided stem, yet the number of tentacles on each head is more than half the average number. This is connected apparently with the fact that the circumference of each half is greater than half the circumference of the original stem. Planarians with double tails, produced by partial splitting, have been described by Dugès and by Faraday, and it has also been shown that by partial splitting of the anterior end of the worm two heads can be produced. Van Duyne, Randolph, and Bardeen and I have obtained the same result. Each half completes itself on the cut-side and produces a symmetrical anterior end. If one of the heads is cut off, it will be again regenerated. If the heads are united very near to the trunk, as in Fig. 42, _A_, they may never grow to the full size of the original head, as I have found; but if the pieces have been split posteriorly, so that each head has a long anterior end, then each one may become nearly as large as the original head (Fig. 42, _B_). We see in these cases the influence of the region of union on the growth of the new part. If the new part is near the region of attachment, the smaller size of the latter restrains the growth of the new head; but if the region of union is farther distant, the head may grow more nearly to its full size despite the influence of the region of union. King has found in the starfish that if the arm is split lengthwise, each half may complete itself laterally and a forked arm result. An additional entire arm may be formed by splitting the disk partially in two between two arms. If the cut-edges do not reunite a new arm will grow out from each cut-surface (Fig. 38, _E_). In this case the development of the new arm cannot be accounted for on the assumption that the typical form completes itself, since a sixth arm cannot be supposed to be a typical structure in the starfish. The result must depend on other factors, such as the presence of an open surface in a region where the cells have the power of making new arms. Barfurth has been able to produce a double tail in the tadpole by the following method: A hot needle is thrust into one side of the tail, so that the notochord and the nervous system are injured. The tail is then cut off just posterior to the region injured by the needle. A new tail grows out from the cut-end, and also in some cases another tail grows out at the side where the notochord was injured by the needle. The injury to the notochord and the removal of tissue immediately about it leads to a proliferation of cells, around which other tissues are added and the new tail produced.