This closing in of the cut-surface, while a preliminary step in the process of regeneration, cannot, I think, be regarded as a part of the regeneration in a strict sense. That the two processes are not dependent on the same internal factors is shown by the following experiments: If a bunch of tubularia is kept in an aquarium, it will produce new heads two or three times and then cease, and if after the last-formed heads have died, pieces of the stem are cut off, they close as readily as do pieces from fresh hydroids. Moreover, at certain times of year the species _Tubularia (Parypha) crocea_ lose their heads, and only the stalks remain. Pieces of these stalks will not regenerate new heads, at this time, although they close in as quickly as do pieces at other times of the year when the heads are present and when new ones regenerate. Another equally good illustration of what seems to be the same phenomenon is found in the closing in of wounded surfaces in the young tadpole embryos. If embryos are taken from the jelly membranes, or even after they have been set free, and cut in half, each piece quickly covers over the wounded surface by means of the ectodermal cells. A much more striking illustration of this closing over in the young tadpole is obtained by cutting, with a pair of small scissors, a large piece from the side. The area may be a fourth or more of the entire side, and yet it may be closed over in an amazingly short time. Half an hour or an hour often suffices to cover a large exposed surface. In this case also the wound is covered not by individual cells wandering over the exposed surface, but by a steady advance of the smooth edge of the ectoderm toward a central point. The process is so similar to that which takes place in tubularia that little doubt can remain as to the two being due to the same factors. As there are no muscle fibres present in the part of the frog’s embryo from which the piece is cut off, the result cannot be due to muscular contraction, but appears to be a contractile phenomenon similar to that in tubularia. Even the small piece that is cut from the side of the body shows the same phenomenon. At first it suddenly bends outwards owing to some physical difference between the inner and the outer parts of the piece. Then the edges thicken, bend in, and begin their advance over the inner tissues. The process is seldom completed, since there appears to be a limit to which the ectoderm can be stretched as the edges advance. A most striking phenomenon both in pieces of tubularia and of the frog’s embryo is the entire absence of dead material at the wounded surface. No sooner is the operation performed than the advance begins; and there is not a trace of dying cells or parts of cells to be seen. CHAPTER IV REGENERATION IN PLANTS