76 



SEVENTH REPORT. 



found to extend almost throughout the length of the column wall and 

 oesophagus. Its appearance is represented in the accompanying figure. 

 The section was taken from the column wall and stained in iron 

 haematoxylin. The nerve layer (ner. 1.) is made up of nerve fibres which 

 interlace a short distance from the mesogloea, and among the fibres are 

 ganglion cells scattered here and there. Also in 1900 J. Stanley Gard- 

 iner (Willey Zool. Results, IV), in an account of the ''Anatomy of a 

 supposed new species of Coenopsammia from Lifu," alludes to and de- 

 picts ectodermal nerve granules, but attaches n© morphological impor- 

 tance to them. 



>^c/6-.' yC- 



hongilxidinal section through column wall of Coenopsammia, eel. ectoderm; end. endoderm; mes 

 mesogloea; mus.l. muscle layer; ner.l. nerve layer. 



In the presence of this well defined nerve layer on its column wall, the 

 coral Coenopsammia differs from all other Madreporaria hitherto de- 

 scribed, but its other anatomical characteristics show the genus to be 

 in no ways exceptional. One would not be prepared to accord a mere 

 special development of the nervous elements any taxonomic importance 

 even among the group of the corals, much less can the genus be regarded 

 as in any way closely related to those Actiniaria in which a similar nerve 

 layer is conspicuous. 



In this connection attention maj" be drawn to a valuable paper on the 

 nervous system of the Actiniaria by Havet (La Cellule, Tom. XVIII, 

 1901) . He was the first to apply with success the Golgi method in tracing 



