298 HARRY BEAL TORREY. 



seg. 2 showing faint signs of tentacles distally; seg. 3 in poor 



shape, removed. 

 April 10, 9.00 A. M. (a) First case. See Fig. 18, b. Note 



tentacles appearing just below ligature. Second case. Similar 



to first case; tentacles on seg. 3 not quite so far advanced. 

 April ii, 2.15 P. M. (a) Segments in both cases broken apart; 



exp. abandoned. 



Of the six cases considered in this experiment, three show 

 development of hydranths immediately below the ligature. 



In a third experiment, the hydranth was removed from a polyp 

 and the column ligatured near the cut and near the base. In 

 three days, tentacles had budded just proximal to the distal 

 ligature, as well as on the small distal segment. The next day, 

 the latter was loosely joined to the segment proximal to it; both 

 segments possessed hydranths with both sets of tentacles. 



These results show that hydranths form readily immediately 

 below the ligature in experiments like the foregoing. 



VI. 



The failure of Tubularia in such experiments to form a hy- 

 dranth immediately below the ligature with the facility exhibited 

 by Corymorpha is correlated with an important structural dif- 

 ference. The stem of Tubularia is encased in a stiff, chitinous 

 layer of perisarc, that offers a certain barrier to the diffusion of 

 gases between coenosarc and the surrounding medium. The 

 column of Corymorpha, furnished with a thin, rudimentary peri- 

 sarc about its base, is naked for more than half its distal length; 

 in this naked distal region, the ligatures w r ere in all cases located. 



When a ligature is passed tightly around a stem of Tubularia, 

 the coenosarc is not only ruptured, but the perisarc, itself intact 

 in most cases, is drawn closely about each end of the coenosarc 

 thus produced. The result is that, while discontinuity has been 

 established, the ccenosarc remains, as before, separated by the 

 perisarc from the surrounding medium. When a ligature is 

 passed tightly around a column of Corymorpha^ discontinuity is 

 established without in any way interfering in a comparable de- 

 gree, if at all, with the diffusion of gases between ccenosarc and 

 sea water. Now, when it is remembered that a discontinuity 



