312 DR. J. E. DUERDEN ON THE 



this respect the irregular endodermal thickenings on the mesenteries immediately behind 

 the simple filaments. The cells are much vacuolated, and here and there Zooxanthellae 

 may occur. To this tissue {/.s., fig. 18), which is partly developed on both the middle 

 and latei'al lobes of the filament, and separates the glandular streak from the ciliated 

 bands, I have already (1900, p. 14G) given the name of intermediate streak. It is 

 recognizable in most trilobed Actinian filaments, and McMurrich (1899, p. 259), in 

 describing the filaments of Zoanthus sociatus, has also independently applied the desig- 

 nation intermediate epithelium to what appears to be the corresponding tissue in the 

 Zoanthean filament. In the pi'esent species it often appears as if folding over or partly 

 covering the first portion of the ciliated bands, and is more trvily endodermal in character 

 than in most Actiniae. In some longitudinal sections through the filaments it presents 

 a distinctly banded appearance, with the deeply staining tissue of the ciliated bands at 

 the bottom of the depressions, as is so well displayed in McMurrich's figures oi Z. sociatus 

 (1899, p. 259). 



The ciliated bands are strongly developed and nearly encircle the lateral mesogloeal 

 lobes. They are constituted altogether of ciliated supporting cells. On the front face of 

 the lateral mesogloeal axes they are continued in many transverse sections as a very 

 narrow band towards the middle ot the filament, diminisliing all the way, and overlaid 

 by the intermediate epithelium, as represented in fig. 18. In other sections the prolon- 

 gations are not recognizable. Posteriorly the ciliated bands pass into what I have 

 termed the reticular streak, which here closely resembles the mesenterial epithelium and 

 is continuous with it. 



In the glandular, intermediate, and ciliated areas the trilobed mesenterial filament 

 thus presents three wholly distinct varieties of cellular tissue. That constituting the 

 glandular streak bears the closest resemblance to the stomodaeal ectoderm ; the inter- 

 mediate tissue is in no way distinguishable from the endodermal epithelium ; while that 

 of the ciliated bands, consisting wholly of long narrow supporting cells, is very specialized 

 and diff'erent from any other of the polypal tissues. 



Prof. McMurrich in his study of the mesenterial filament of Z. sociatus has devoted 

 .si^ecial attention to the relationship of the ciliated bands and the intermediate region. 

 With its long, well-developed filaments, the species is specially fitted for such an investi- 

 gation. Discussing the endodermal origin or otherwise of the intermediate epithelium, 

 McMurrich affirms that it is certainly continuoiis with the stomodaeal ectoderm above, 

 and not with the endoderm, and from all the evidence before him is inclined to regard it 

 as being ectodermal like the rest of the ciliated baud epithelium. So far as histological 

 evidence can be dejjended upon, an endodermal origin would be assigned it in the 

 present species ; in character it is wholly distinct from the ciliated and glandular streaks 

 and stomodaeal ectoderm. 



The trilobed condition of the mesenterial filaments in Bunodeopsis is continued only for 

 a short distance down the mesenteries, when it becomes replaced by the simple filament ; 

 but the species is not suitable for tracing the actual passage from the one to the other. 

 McMurrich (p. 2G1) states " that in Z. sociatus there is neither a histological continuity 



