Studies on Regulation. 491 



It is difficult to understand otherwise why the margins of this 

 region are concave instead of convex or straight. 



But the form differs according to the level from which regenera- 

 tion occurs (compare Figs. 11, 15 and 16, 24 and 25). The 

 tapering outline with concave margins is most marked when 

 regeneration takes place from a level near the anterior end. This 

 difference is of course connected with the fact that the amount or 

 regeneration is greater in anterior regions but that difference does 

 not explain why the new tail should be more slender in the one 

 case than in the other. Evidently this difference in form of the 

 new part at different levels is mechanical. In the first place, 

 when the regenerating part represents a large part or all of the 

 region of the body used for attachment it performs the functions 

 of that part and is subjected to the tensions resulting from this 

 function. When, however, as in Figs. 24 and 25, it represents 

 only the posterior region of the part used for attachment the lateral 

 margins of the old portion anterior to it perform in large degree 

 the function of attachment and hence the regenerating part is sub- 

 jected only to a relatively slig-ht degree of tension. Therefore, it 

 is less elongated, less concave laterally and more blunt posteriorly. 



In the pieces without ganglia these differences are naturally 

 much less marked since the tension upon the parts resulting from 

 locomotion is relatively slight in all cases. But even here there is 

 a difference, at least between Series 73 and the others (compare 

 Figs. 9 and 10, Series J^, with Figs. 17 and 18, Series 79, and 

 Figs. 26 and 27, Series 81). The peculiar form of the regenerating 

 part in Fig. 9 seems to be due to the fact that this piece used one 

 part of the margin as often as another for attachment, whereas 

 usually the median posterior region is used more than other parts. 

 The form of the regenerating part shown in this figure is common 

 in such pieces and may, I think, be regarded as due primarily to 

 lack of coordination; any part of the margin which happens to 

 come into close contact with the substratum or which is stimulated 

 in any other way becomes attached and consequently the part 

 does not taper posteriorly. The piece shown in Fig. 10 was 

 able to progress more rapidly and in general showed a higher 

 degree of coordination than the other. The difference in the 



