i6 Gcor<:c TJioriias Hars[iti 



i'V." 



Proximal Tentacles. — In the formation of the proximal 

 tentacles, the complex folding of T. crocca does not take place, 

 nor is the very evident evagination of T. tcncUa the method, 

 but rather a combination of the two. The development begins 

 very much the same as in T. tcncUa (figs. 34, 35) by what seems 

 to be an evagination of the entoderm. A sort of ridge is thus 

 formed. The ectoderm folds around the entodermal cells (be- 

 ginning at the distal end), very much the same as in T. crocca. 

 By a continuation of the evagination and folding, the entoderm 

 cells are surrounded by the ectoderm and the tentacle cut off 

 (figs. 32, 33). The tentacle when thus cut oft' is of normal 

 length. Unlike T. crocca, the gonads begin their formation before 

 the hydranth emerges from the perisarc (cf. figs. 13a, 13&). 



Rather strange is the occurrence of nematocysts in the ento- 

 derm and in the debris of the enteric cavity. This is most com- 

 mon in the regenerating stems of T. larynx, though occasionally 

 found in other forms. Weismann (1883) explains the presence 

 of nettling cells in the developing male gonad of Ciava squaiuafa 

 as a protection against a parasitic fungus. He performed a num- 

 ber of experiments with the result that the male form could not 

 be infected, while the female gonad, whicli was not provided with 

 the nettling cells, was easily infected with the fungus. Such an 

 explanation could scarcely account satisfactorily for the pres- 

 ence of the nettling cells in the entoderm. This explanation would 

 prove to be especially inadequate since in other forms, in which 

 the nettling cells are not found in the entoderm, the open end 

 closes no c|uicker than in T. larynx. 



Among the colonies of 7\ larynx ol:)tained during the summer, 

 three pieces were found to be naturally regenerating hydranths. 

 These were killed and later sectioned. The folds in the hydranth 

 region were extremely complex, often pinnate, and entirely unlike 

 anything observed in the artificially regenerating hydranths of T. 

 crocca, T. tcncUa, or T. larynx. Owing to the very limited amount 

 of this material no conclusions can be drawn, and no explanation 

 offered, unless it be possible that su.ch regeneration was the re- 

 sult of abnormal conditions, though it is not easy to conceive 

 what they might have been. 



290 



