52 



subsequently there is likewise no description bearing on the subject. In 

 some instances, at least, this has been due to the fact that this organ 

 has been dislodged, owing possibly to rough handling in the trawl or 

 dredge or to violent muscular contractions on the way to the surface. 

 In two of the specimens in hand the proboscis is intact, and in one case 

 is almost completely extended. This last named example was carefully 

 cleared, and subsequently the proboscis was completely sectioned, but 

 there is absolutely no trace of any stylet. It is possible that such a 

 structure may have disappeared, owing to decomposition of the formal- 

 dehyde, but this would not remove the muscular or connective tissue to 

 which it is attached in other metanemerteans. From first to last the 

 proboscis is a circular tube, without special modifications. 



In cross sections the proboscis conforms closely to the metane- 

 mertean type. The external epithelial covering is faintly distinct here 

 and there as an excessively thin sheet, but its finer details are not suf- 

 ficiently marked to warrant a description. Beneath this layer is the 

 longitudinal set of muscles separated into two divisions by the ner- 

 vous layer. The outermost division stains more darkly than those of 

 the outer division, but otherwise there appears to be no essential dif- 

 ference between them. The proboscidial nerves, numbering approxi- 

 mately 22, are fairly well defined in certain localities, and the fibers 

 passing from them form the usual plexus. The inner layer of circular 

 muscles is well defined in the forward division of the proboscis, but in 

 the succeeding glandular section it is absent or reduced to a very few 

 fibers more or less buried in the connective tissue stroma that supports 

 the lining, glandular epithelium. This last named layer has been dis- 

 lodged in most places, and the component cells are accordingly ill- 

 defined. They agree however in being slender elements, with sub- 

 central nuclei, and distally contain quantities of a darkly staining 

 secretion after treatment with Delafield's haematoxylin. The lining 

 epithelium of the anterior, non-glandular section of the proboscis con- 

 sists of a thin sheet of cells, without distinct boundaries, resembling 

 closely those bounding the outer proboscidial surface. 



As in other species of the genus the digestive tract conforms to the 

 metanemertean type. The mouth is distinct from the proboscis, but in 

 the position of the stomach, pylorus and intestine with its anterior coe- 

 cum the resemblance is decided. The mouth cavity is comparatively 

 large, though its highly folded walls obliterate the lumen to some ex- 

 tent, while the oesophagus is extremely limited in length. Posteriorly 

 this leads into a stomach of slender outline that about opposite the level 

 of the posterior border of the cirri unites with the intestine. This last 

 named organ bears a close resemblance to what exists in N. pelagica, 



