12 THE VOYAGE OF H.M.S. CHALLENGER. 



tentacles ; besides the marginal, one oral tentacle and one intermediary are placed 

 on the disc. In the intra-mesenterial chamber of the second order the former (oral) was 

 wanting; in those of the third order, the intermediary tentacle was also wanting, or 

 rather was replaced by a peripheral tentacle. From all the inter-mesenterial chambers, 

 and also from the supernumerary intra-mesenterial chamber, spring only marginal 

 tentacles. 



Histologically our Corynactis is closely related to the Corallimorphi. The 

 mesoglcea is homogeneous, branched stellate cells are richly scattered in it, while the 

 modified bladder cells, which occur in Corallimorphus obtectus, are wanting. Beneath 

 the endoderm runs a fibrous layer, sometimes closely under it, sometimes separated from 

 it by a homogeneous layer, giving oif bundles which run to the endodermal surface. 

 The musculature of the oral disc and tentacles is weak and ectodermal ; there is no 

 special sphincter, and the mesenteries are provided with only weak muscles on both sides. 

 On the other hand, I was surprised at the occurrence of longitudinal muscles on the 

 outer side of the hody-ivcdl They are not very strong, and are mostly composed of 

 short spindle-shaped fibres, the lamella being always slightly pleated here and there. 

 This discovery made it necessary to study Corallimorphus obtectus anew, with reference 

 to the body-wall. The epithelium having been preserved only at exceedingly few spots, 

 constituted the reason why I had not previously observed the muscle, but a renewed 

 study yielded figures by which I arrived at the following definite opinion, based on 

 numerous preparations from difi'erent parts of the body. 



At the basis of each epithelial cell lies a small body, staining in carmine, and 

 resembling, in sections accurately transverse to it, a muscle fibril. If the section be 

 taken at an angle of about 30°, these bodies appear elongated and somewhat spindle- 

 shaped ; but I have seen no such obvious longitudinal fibres as in Corynactis. I am 

 therefore of opinion that Corallimorphus possesses longitudinal muscles, but that they 

 are extremely rudimentary. 



The observation of ectodermal longitudinal muscles on the body-wall of Corynactis, 

 is ijcv exceptionally interesting discovery. Among all Anthozoa, we know of a similar 

 condition in Cerianthus alone, and, as I may here mention, anticipating future 

 investigation, in Arachnactis, a genus very closely allied to Cerianthus : while in 

 the typical Anthozoa the ectodermal musculature is confined to the tentacles, the oral 

 disc, and the stomatodseum. On the other hand, all Hydroids in the hydra-form {i.e. 

 Hydroid-polypes and Scyphostomse) possess ectodermal longitudinal muscles of the 

 body-wall, which are prolonged directly into the tentacles and oral disc (peristome). 

 We have here, throughout the whole body, circular muscles on the endodermal side, and 

 longitudinal, i.e. radial, on the ectodermal. 



On the ground of previous researches on the sexual organs, I have published the 

 view, since defended by Gotte, that the Scyphomedusse are ancestral forms of the 



