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



folded in zig-zags, the next and longest region (cEsophagus) is mostly devoid of tlicm, but 

 they again become very well marked in front of the dilatation (stomach proper) generally 

 found at the posterior end of the body. The intestinal canal then bends forward along 

 the neural aspect to terminate in the anus — on the elevation between the two branchial 

 fans. The oesophageal and stomachal regions are fixed by a median ventral mesentery 

 and the various radiate muscles formerly described, and by the two intestinal 

 mesenteries on its posterior face ; one of these mesenteries (the right) and a median 

 dorsal, moreover, fixing the intestine in a special manner. 



The mucous surface of the canal is composed of a cylindrical cellular layer some- 

 what resembling the hypoderm. It forms a thick and richly folded granular and 

 streaked tissue in the region at the base of the branchial fans. In the narrow 

 pharjmgeal division the layer attains considerable thickness, the folds forming numerous 

 areolae when viewed from the surface. This condition is less marked in the oesophageal 

 region. In what may be termed the first stomachal region immediately behind the 

 former, the streaked glandular tissue has a finely cellular and granular aspect on the 

 surface. In the dilated stomachal region proper, the folds of this tissue again increase 

 in thickness. The mucous coat of the intestine is similar in structure, but much 

 thinner. 



Towards its termination the rectal portion of the gut generally retains a somewhat 

 triangular outline in the preparations. On approaching the level of the constricted 

 canals of the nephridia, the basement-layer of the rectum shows externally and posteriorly 

 crenate processes ; then a kind of reticulation occurs in the same tissue with muscular 

 fibres (PI. II. fig. 1), and the canal terminates in a small circular aperture, which still 

 has a ring of basement-tissue beneath the mucous lining. The terminal part is thus 

 very muscular as well as elastic, and well fitted to send a jet of faeces a considerable 

 distance. Dyster, indeed, observed the latter in the living form voided by jerks, the 

 fusiform pellets being connected by slender filaments. These fusiform pellets are 

 common in Appendicularia, which feeds on similar food. 



Food seldom occurs in the first regions of the alimentary canal, but in the 

 stomachal and intestinal portions the granular contents abound in Foraminifera, 

 Radiolarians, Diatoms, spicules of Echinoderms and sponges, with other organic debris. 

 Very little mud and few or no sand-grains occur in the intestine, the organisms just noted 

 forming elongated and coherent masses by aid of a translucent and finely granular 

 stroma, probably the result of secretion, though in some the great abundance of minute 

 greenish granules suggested the possible presence of such low gelatinous organisms as 

 occasionally occur in our own seas. 



In the larval form (Actinotrocha) E. Wagener found Bacillaria, Peridiniae, and 

 spores of Algae, and Schneider mentions similar forms with Diatoms in the same stage ; 

 moreover, in the Sipunculoid form into which the larva is metamorphosed, he describes 



