8 George Thomas Hargitt 



change or development took place although the pieces were kept 

 two weeks. Several other like experiments were tried, with both 

 T. cvoeea and T. teiiella. In no case, however, did any growth 

 take pace nor any sign of regeneration occur, except the col- 

 lection of red pigment. This latter condition was found in per- 

 haps four or five pieces out of one hundred or more. 



Experiment 24. — Tubularia eroeea cut into pieces 12 to 20 

 mm. in length and placed on the 1)ottom of a dish. Some were 

 cut obliquely, some scjuare across. In 24 hours the tentacle an- 

 lagen had developed in several pieces (fig. 3) ; 36 hours later 

 the anlagen had been formed in twelve more pieces, and 24 

 hours after this ten new pieces showed the anlagen. After this 

 the anlagen developed in the pieces gradually, so that in ten 

 days all but five pieces out of fifty showed them. The proximal 

 series seemed to be developed first. The general features of 

 the development of the hydranths have been considered by other 

 authors. First the coenosarc withdraws and the cut edges close 

 in. A\^hether this closing in is a simple bending over of cut 

 edges (Morgan, 1901), or due to amoeboid motion of the whole 

 membrane (Morgan, 1902), or to a centripetal force acting on 

 the cells at the cut edges (Stevens, 1902), I did not determine. 

 After this membrane has been formed a thin perisarc is se- 

 creted over the open ends and the circulation of the enteric fluid 

 begins, the longitudinal ridges of entoderm disappearing in the 

 hydranth region. Then the tentacle anlagen are folded off, ap- 

 pearing as ridges. This folding seems to take place along the 

 entire length of the tentacle, beginning at the distal end. The 

 proximal anlagen are formed first. All this has taken place 

 within the perisarc and seems to be a transformation of tissue 

 as stated by Bickford (1894) for T. feiiella. When the analgen 

 have been completely formed, the thin perisarc over the end of 

 the piece is ruptured, the coenosarc elongates, the hydranth is 

 forced out and assumes the perfect hydranth form. The gonads 

 ^vere never noted as developing till later. In several pieces this 

 process was followed only to the formation of tentacle anlagen ; 

 then the coenosarc was pushed out of the, perisarc and the ten- 

 tacles appeared as small knobs or buds, attaining their full size 

 by further growth. 



282 



