DYNAMICS OF MORPHOGENESIS 73 



ment of a new axial gradient and so the reorganization of the 

 region y. 



As regards the region z of figure 2 from which the posterior 

 end develops, the relations to other parts are not as clear as in 

 the case of x, for the chief visible differences in tail-formation are 

 merely differences in amount of growth. In general, however, 

 it is evident that development of the posterior end is directly 

 proportional to rate y. But the stimulation following section 

 has no relation to tail formation, because it is merely temporary 

 and the tail is a subordinate part, depending for its very exist- 

 ence upon correlations with more anterior regions. It can form 

 at any time, whenever rate y is high enough to determine its 

 development. Tail-formation is retarded or inhibited by all 

 depressing conditions, such as low temperature, low concentra- 

 tions of narcotics (Child '12), etc., but exactly the same con- 

 ditions may increase head-frequency in the same pieces. The 

 experimental data upon this point will be presented at another 

 time. 



IV. THE FACTORS WHICH DETERMINE LOCALIZATION OF THE 



NEW HEAD 



In all longer pieces of Planaria, the head when it forms is local- 

 ized at the anterior end of the piece. The basis for this locali- 

 zation is the axial gradient: how the gradient determines the 

 localization we have now to consider. Attention has already been 

 called to the fact that when a planarian (earthworm, etc.) is cut 

 in two, the posterior piece is much more strongly stimulated than 

 the anterior. The same relation is evident in contact stimu- 

 lation. SHght stimulation of a given region produces much more 

 marked effects posterior than anterior to it. In short, the 

 whole mechanism of dynamic correlation in Planaria and similar 

 forms is developed on the basis of the axial gradient. Correla- 

 tion between regions of the body is chiefly in the posterior direc- 

 tion, anterior regions being relatively independent of posterior, 

 and posterior regions relatively dependent upon anterior. That 

 metabolic gradients exist in at least some nerves in the lower 

 animals is known (Child '14 a) and there are reasons for be- 



