FORM REGULATION IN CERIANTHUS. 283 



those which are collapsed. It is possible to observe on the liv- 







ing specimens that the tentacle regeneration is more rapid in dis- 

 tended than in collapsed portions. Though it may not always 

 be possible to determine exactly the causes that have operated in 

 bringing about a particular case of local collapse or compression 

 its presence can be recognized without difficulty, even before ten- 

 tacle regeneration has begun. 



A few examples selected from the numerous cases of this kind 

 observed will serve to illustrate the subject. Fig. 5 represents 

 a case in which one side of the oral region of a piece became 

 folded inward during collapse and so failed to become distended 

 when closure was allowed to occur. It will be observed that the 

 new tissue closing the oral end does not form a rounded area but 

 simply fills the crevices between the inrolled margins. The fail- 

 ure of the new tissue to distend is due to the fact that the aboral 

 end was kept open until the new tissue had become thick 



FIG. 5. FIG. 6. 



enough to resist the tension and to grow very slowly. It will also 

 be observed that with approach to the infolded region the mar- 

 ginal tentacles decrease in size and finally cease. The region 

 without tentacles represents a much greater portion of the cir- 

 cumference than is apparent in its contracted condition. The fig- 

 ure is simply a view of the oral end. A short distance from the 

 end the fold disappears and the body is cylindrical in form. The 

 fold was gradually obliterated in this piece and the missing ten- 

 tacles finally appeared. 



In Fig. 6 a similar case is shown, except that here two folds 

 are present on opposite sides of the body. In the region of each 

 of these folds the tentacles diminish in size and disappear as in 

 the case of Fig. 5. 



In these cases it is of course impossible to determine just why 

 the water has failed to enter the infolded portions, but it is prob- 



