EEPORT ON CEPHALODISCUS DODECALOPHUS. 41 



The course of the alimentary canal is sufficiently explained by means of fig. 2.' 

 The mouth (m.) is seen to be overhung by the large proboscis (p.), otherwise known as 

 the epistome or buccal shield. The dorsal and ventral mesenteries are not visible, since 

 the section is not exactly median. The third body-cavity (b.c.^) is very large, and contains 

 the alimentary canal and ovaries. The anterior body-cavity (b.c.^) continues to form the 

 lumen of the proboscis, whilst the relations of the collar-cavities are not quite those of 

 the preceding figure. Dorsally the two halves of the ' body-cavity of the collar have 

 extended forwards into the proboscis region, whilst ventrally they are to be found 

 (appaxently fused ?) solely in the post-oral lamella or operculum. The cavities of the 

 proboscis and of the collar (and more particularly of the latter) are to some extent 

 obliterated by muscles and connective-tissue. The notochord (nch.) has, in the adult, 

 the form of a slender bar, provided with a fine lumen, stretching forwards in the proboscis- 

 stalk into the proboscis itself. It is continuous at its base, as in the young bud, with 

 the epithelium of the alimentary canal. 



The further relations of the notochord will be described in connection with fig. 4. 



The central nervous system is developed on the dorsal side of the collar as a mass of 

 ganglion-cells and netve-fibres lying outside the basement-membrane of the epidermis. 

 It is, however, continuous anteriorly with a similar development of nervous tissue 

 situated on the dorsal' aspect of the proboscis, and laterally with a well-developed ner\^e- 

 layer on the dorsal sides of the lophophoral arms. In the young bud, in which the 

 collar is represented by a region of the body separated by transverse gi-ooves from the 

 proboscis on the one hand and the trunk on the other, the condition of the central 

 nervous system as a development of the collar is particularly well marketl, although even 

 at an early stage a thinner nerve-layer occurs on the proboscis. In the adult, there is no 

 sharp line between the nervous tissue of the collar and that of the proboscis. At about 

 the level of the anterior end of the notochord, the nerve-layer is perforated by a pair of 

 pores — apparently derivatives of the ectoderm, which pass from the exterior into the 

 body-cavity of the proboscis {p.p., fig. 2). These pores are disposed symmetrically with 

 regard to the median plane of the animal, and are at no great distance frLim rlmt 

 plane. 



In most species of Balaiioglossus, it is well known that an asymmetrical proboscis- 

 pore occurs on the left side, whilst in Bala7}oglossus hipfferi '' two proboscis-pores are 

 present. It is thus obvious that the proboscis-pores of C'e2)hulodiscus arc a furtlier feature 

 in support of the \aew that this animal is related to Balanoglossus, the disposition of 

 these pores in Balanoglossus hipfferi being in this respect particularly noteworthy. 



One of the two ovaries [ov.) is represented in section in fig. 2. The duct (ore?.) is 



'A f'.Ul in the inteatinal region of the alimentary canal, of the existence of which I have been informed by 

 Professor M'Intosh, has not Ixun inserted in the diagram. 



' Vide Bate.sun, Quart. Journ. Micr. Sci., vol. xxvi. p. 550. 



(ZOOL ClIALL. EXF. — I'ABT LXII. — 1887.) Ql1 ** 



