REPORT ON THE ANTIPATHARIA. 217 



almost invariably collected into batteries, as in many Madreporaria. Such an arrange- 

 ment of the nematocysts is not usual in the Actiniaria, but is found in the marginal 

 spherules of certain forms. The batteries of nematocysts are usually surrounded 

 by a more or less complete ring of gland cells. The gland cells may be either hyaline 

 or granular, but the two types are apparently not found in the same species. A 

 nervous layer connected with a row of slender fibres, which are probably sensory in 

 function, is always present near the base of the ectoderm, and ganglia have been demon- 

 strated in it in many instances. An ectodermal muscular layer is apparently always 

 present, but has a variable development. In some genera it is quite unimportant, whereas 

 in others {e.g., Leiopathes and Cirripathes) the fibres are of considerable thickness 

 and are applied to the dentate surface of the ruesogloea, so that the layer is then some- 

 what convoluted. The ciliated epithelial cells (Stiitzzellen) do not form so important a 

 feature of the ectoderm as is the case in Hexactinise. The ectoderm of the stomodaeum 

 is, so far as my observations go, entirely devoid of nematocysts. It contains granular 

 gland cells of variable size and usually also a number of the hyaline type, which are 

 situated near the base of the layer. The " Stiitzzellen " are here more important, and are 

 separated from one another by elongate fibrous cells. Unlike the Hexactinise the stomodseal 

 ectoderm of Antipathinse frequently has a muscular layer at its base. A similar layer is 

 also generally present on the body-wall beneath the insertion of the tentacles, in that 

 part namely, where, according to the researches of the Hertwigs, ectodermal muscular 

 fibres are absent in Hexactinise. 



The mesoglcea has a similar structure in all the Antipathinas examined. It consists 

 of a hyaline or subfibrous layer of variable thickness, and has never been observed to 

 contain isolated connective-tissue cells as in Hexactinise. In Cerianthidse these rounded 

 or stellate connective-tissue cells are, however, rare. In Cladopathes, amongst the 

 Schizopathinse, the mesoglcea is relatively very thick, and is furnished with isolated 

 stellate cells of the Actinian type. 



The entoderm usually contains only the hyaline type of gland cells. Apparently the 

 glandular elements are more numerous than in Hexactinise, and the Stiitzzellen are less 

 so. The surface of the entoderm frequently consists of an irregular cubical epithelium 

 similar to that of many Madreporaria. No nematocysts have ever been observed in any 

 part of the entoderm. The nervous layer is apparently always represented, but the 

 entodermal muscular system may be rudimentary or absent. In most types the inner 

 surface of the mesoglcea is flat, and a more or less important layer of entodermal muscular 

 fibres may be applied to it, which appears first recognisable in vertical sections of the stomo- 

 dseum. In Leiopathes and Cirripathes the inner (as well as the outer) margin of the meso- 

 glcea is dentate, and bears a more or less convoluted layer of entodermal muscular fibres. 

 These are the only genera in which an approach to the entodermal muscular system of 

 Hexactinise has been observed. No muscular fibres have as yet been observed in the 



(ZOOL. CHALL. EXP. — PART LSXX. — 1889.) Llll 28 



