4/6 C. M. CHILD. 



ties resulting from lateral injury are not similar to the lateral 

 partial discs which result from a transverse cut part way through 

 the body in Cerianthus (Loeb, '91, Child, '05, '08) and Harenactis 

 (Child, '093). In those forms the opening remains because the 

 cut edges of body wall and oesophagus unite and the new partial 

 disc develops entirely on the proximal side of the cut, just as a 

 disc develops on any distal cut end of a piece. In the case of 

 Coryinorpha the wound closes completely in the course of a few 

 hours and it is only through the continued growth of the region 

 after closure of the wound that the new axis is determined. If 

 the injury does not initiate such growth no new axis develops. 



These new lateral polarities are essentially induced buds and 

 like other buds they give us important evidence concerning the 

 origin and nature of new axes. If we observe, without theoretical 

 prejudice, what happens in such a process, we see that the new axis 

 originates as a local region of growth and becomes visible as an 

 outgrowth of the body wall (Fig. 2) because the growth activity 

 is evidently greatest in its middle region and decreases peripher- 

 ally in all directions. The early rounded outgrowth undergoes 

 elongation and the more active middle region necessarily becomes 

 its tip (Fig. 3), in other words, the region of growth has now be- 

 come a physiological axis characterized by a gradient in activity 

 decreasing from the tip basipetally. There can be no doubt that 

 when such a gradient is once determined in a particular kind of 

 protoplasm the constitution of the protoplasm will play the chief 

 part in determining its steepness, its length and the changes which 

 it undergoes during development. If we admit this, it follows 

 that however the gradient is determined its definitive form will 

 be the same in a protoplasm of a certain constitution, consequently 

 a gradient such as the one under consideration, determined by a 

 local injury will determine the same course of development, i.e., 

 the same kind of an axis as the gradient in embryonic development, 

 if the condition of the cells in the region of active growth is similar 

 to that of embryonic cells of Cormorpha. If we take the facts 

 as they stand it seems that there is no adequate reason to regard 

 a polar axis in its simplest form as anything more than such a 

 gradient as this. Experiment shows that when such a gradient 

 is determined a new axis is determined and when the gradient is 



