256 McMURRICH [Vol. II. 



with green granules and fragments of sponge spicules. I sug- 

 gested that this region of the endoderm was essentially digestive 

 in function, an opinion which has since been confirmed experi- 

 mentally by Willem ('93) for the Hexactiniae. 



Haddon and Shackleton ('91 and '9ia) confirmed my observa- 

 tions on other zoanthids and pointed out that slight variations 

 in the form of the ciliated lobes occurred in certain forms, such 

 as P. axmellae. Finally, von Heider ('95), in his description of 

 Z. chierchiac, entered somewhat fully into the structure of its 

 mesenterial filaments. He found in the ciliated bands what he 

 regarded as a distinct area intervening between the central lobe 

 and the ciliated portions of the lateral wings, and characterized 

 by possessing numerous gland cells. He terms it the glandular 

 swelling and attributes to it a digestive function. He also 

 observed the heightened epithelium lateral to the glandular 

 portions of the filaments which had been described by Haddon 

 and Shackleton, and myself, but objects to its interpretation as 

 a digestive area, believing it to be the region in which the 

 reproductive cells are to develop, none of the specimens which 

 he examined possessing these cells. 



H. Descriptive. 

 I. The Ciliated Bands. 



To correctly interpret a series of transverse sections of an 

 adult Zoanthus it is necessary to understand the course of the 

 free edge of the mesentery. For this purpose I made a wax 

 reconstruction of the upper part of one of the perfect mesen- 

 teries of Z. sociatiis, together with the portion of the stomato- 

 daeum to which the mesentery was attached. From this it is 

 evident that the lower edge of the stomatodaeum is bent back 

 upon itself, as represented in the diagram (Fig. i), and that its 

 ectoderm becomes continuous with the epithelium of the large 

 ciliated bands. From the reflected stomatodaeum the free edge 

 of the mesentery, with the filament, extends outwards and then 

 arches downwards. Consequently, in transverse sections through 

 the column the filament will be cut practically longitudinally 



