374 ON THE ANATOMY OF A SUPPOSED NEW SPECIES 



the exterior under the structureless membrane there is commonly found a thin layer 

 of granules exactly similar in appearance to the granules found at the bases of the 

 overlying endoderm. 



SECTION IV. 

 Conclusions relating to the Bodv Layers in the Actinozoa. 



The central glandular elements of the mesenterial filaments have been shown by 

 the Hertwigs (14) and M'Murrich (22) for the Actiniaria, and E. B. Wilson (27) and 

 Hickson (16) for the Alcyonaria to be the elements, which produce the digestive 

 secretion. Similar scattered glandular elements have been recorded by most observers 

 in the stomodoea of Actiniaria and iladreporaria, and recently hav^e been found also 

 by Ashworth (1) in the stomodoeum of Xenia. The abundance and regular arrangement 

 of the glandular elements, found in this species of Coenopsammia has however not 

 previously been observed. In longitudinal sections through the mesenterial filaments 

 and stomodoeum it is impossible to tell where the latter ends and the former com- 

 mences so complete is the resemblance between the epithelia of these two parts. The 

 epithelium too of the lateral parts, the " Flimmerstreifen " is precisely similar to that 

 of the stomodoeum between the attachments of the mesenteries. It is important also 

 to remember that the mesenterial nematocysts are found on all the three parts of 

 the filaments and also in the stomodoeum and external ectoderm of the body. 



It was suggested first by E. B. ^yilson (27) that the lateral parts or "Flimmer- 

 streifen " of the Actiniarian mesenterial filament are ectodermic in origin and that the 

 median part, or " Drlisenstreifen," is endodermic. Fowler (7) however from the histology 

 of the Madreporarian filaments considered the median part to be ectodermic and the 

 lateral parts to be its unbroken gradation into the endoderm but mainly endodermic. 

 H. V. Wilson (28) then from the development of Manicina came to the conclusion 

 that the whole filament is ectodermic in origin. Indeed he showed that the filaments 

 of the twelve primary mesenteries are formed as outgrowths from the basal end of 

 the stomodoeum. M'^Murrich (22) would not admit the homology of the lateral parts 

 of the Madreporarian filaments with the "Flimmerstreifen" of the Hert\vigs and main- 

 tained the views of E. B. Wilson. 



In this species of Coenopsammia it is obvious that if the central part of the filament 

 is endodermic, a great part of the wall of the stomodoeum is likewise endodermic. 

 Ashworth (1) however states that in Xenia, where digestive cells also occur in the 

 stomodoeum, he has followed the development in the bud and that the stomodoeum 

 is entirely ectodermic. I have also found that the stomodoeum is entirely ectodermic 

 in its origin in the buds of Pocillopora. In this species of Coenopsammia the glandular 

 elements are found right up to the mouth of the stomodoeum and in the bud 

 formation, so far as I could follow it, the whole stomodoeum appears to be formed 

 by the ectoderm. The lateral parts of the mesenterial filaments are similar in structure 

 to the "Flimmerstreifen" of the Hertwigs and have apparently the same function. 



