GEPHYREA. 619 



in other Gephyrea. The digestive tract itself takes a straight course from 

 mouth to anus in Priapulus (except in P. glandifer} and Halicryptus ; it is 

 convoluted in Echiuridae, principally in side to side loops, whereas in the 

 Sipunculidae it is coiled in a close dextral spiral, except in certain tubi- 

 colous forms, in which the coils are loose. The axis of the spiral is gene- 

 rally occupied by a muscle, usually attached in front close to the anus, and 

 frequently behind as well, at the apex of the body. From this muscle 

 fibres pass to the walls of the tract. Muscle-fibres, especially well-developed 

 in Echiuridae, also attach the tract to the body-walls. The pharynx and 

 the rectum are always provided in Gephyrea achaeta with specially deve- 

 loped bands, which attain great prominence as the retractor muscles in 

 connection with the former. The retractors in question are derived from 

 the longitudinal muscle-layer of the body ; are short in the Priapulidae, 

 of considerable but variable length in the Sipunculidae. In the last-named 

 there is either a single ventral retractor, or two ventral, or two ventral and 

 two dorsal. The tract itself is divisible in Sipunculidae into a muscular 

 pharynx, intestine, and extremely short rectum. In Echiuridae it con- 

 sists of an anterior (* buccal ') region, which ends at the spot where the 

 dorsal blood-vessel forms a peri-intestinal ring, of a middle region of great 

 extent, in which the longitudinal muscle-layer is internal to the circular 

 instead of the reverse as is usual, and of a short rectum. In the Echiurids, 

 Bonellia, Thalassema, Echiurus, and possibly Hamingia, a siphon, or col- 

 lateral intestine, comparable to the structure seen in some Echinoids, ex- 

 tends for the greater part of the length of the middle region (supra) on its 

 ventral aspect. A band of longitudinal muscle-fibres underlies the siphon. 

 The inner aspect of the pharynx is armed in Priapulus and Halicryptus 

 with numerous teeth, analogous in structure with the spines of the body; 

 the cuticula lining it is fairly thick, as is that of the rectum, and both alike 

 are shed when the outer cuticle is cast off. The tract is ciliated throughout 

 in Echiuridae, and there is a ventral furrow with especially long cilia in its 

 middle region. So, too, in Sipunculus nudus, according to Keferstein. A 

 delicate cuticle lines the tract in Priapulidae. Thalassema has a ventral 

 caecum at the end of the middle region, which has been found filled with 

 a gelatinous substance in Th. erytkrogramtnon. The rectum in the same 

 'genus is dilated terminally, and a layer of peculiar cells underlies its 

 peritoneal epithelium. A caecum is sometimes present at the commence- 

 ment of the rectum in Sipunculids, and tufts of branched caeca close to the 

 anus. But in Echiuridae with some exceptions (Saccosoma, Epithetosomd}^ 

 there is always a pair of anal caeca, the branchiae of Greefif, or posterior 

 nephridia appended to the rectum and opening into it. They have mus- 

 cular walls, and communicate with the coelome by means of ciliated 

 funnels with short stalks. In Bonellia viridis and Hamingia, they bear 

 a number of branched stems, each branch ending in a funnel. The 



