12 DERMAL SYSTEM. 



consist almost entirely of bodies exactly resembling the nuclei of the cells ; some of these 

 intercellular nuclei contain two nucleoli, and seem to be undergoing division, while round 

 others a young cell may be seen in various stages of formation. 



These different stages may be satisfactorily followed, and afford a very interesting example 

 of cytogenetic action. The first thing observed is the accumulation round the nucleus of a 

 little mass of granular protoplasm. In the midst of this a minute vacuola next shows itself; 

 this becomes the cell-cavity, and gradually increases in size with the enlarging cell ; the 

 nucleus is persistent, remaining attached to the cell-wall. 



For reasons to be presently mentioned, it is highly probable that the endocyst is pervaded 

 by a system of canals of extreme delicacy, which constitute an irregular network in its 

 substance. 



Besides the structure now described, peculiar fibres (PI. II, fig. 10) are also developed 

 in the endocyst. These are situated on its inner surface, where they constitute a well-marked 

 layer composed of a network of transverse and longitudinal fibres over the whole extent of the 

 endocyst. I have even succeeded in separating this network as a conthiuous layer. 



There can be no doubt that these fibres of the endocyst are muscular, and that it is to 

 their presence that the high degree of contractility enjoyed by the endocyst of Lophopim is 

 in a great measure, if not entirely, due. When treated with acetic acid, they are plainly seen 

 to be composed of greatly elongated fusiform cells, having their pointed extremities in con- 

 nection with one another. Each of these cells is then also seen to contain a nucleus with 

 nucleolus (PI. II, figs. 11, 12). 



We have seen that in all the fresh-water genera a portion of the endocyst remains in a 

 permanently invaginated state. It is probable that in all these genera the endocyst retains 

 its general structure and contractility for a greater or less extent of its permanently invaginated 

 portion, down to a spot where the transverse fibres appear to become condensed into a sort 

 of sphincter, and shortly after this the endocyst alters its texture, losing its contractility and 

 becoming thinner (PI. V, fig. 6). In this condition it continues till it terminates by being 

 attached to the base of the tentacular crown. This thin, non-contractile portion of the 

 endocyst constitutes the tentacular sheath which encloses and protects the tentacula during the 

 retracted state of the polypide. Near the spot where the endocyst passes into the tentacular 

 sheath, there appears to exist, at least in Lophopus, a circular canal, which here passes trans- 

 versely round the endocyst. The presence of this canal is revealed by peculiar, spherical or 

 oval, brilliant corpuscles, which it almost always contains in Lophopus. 



A portion, perhaps the whole, of the inner surface of the endocyst is clothed with 

 vibratilecilia. 



Though I have not succeeded in making out the structure of the endocyst in the other 

 genera so satisfactorily as in Lophopus, we may, nevertheless, conclude that it is nowhere 

 very far different from that now described. In all these, fibres may be detected in the endo- 

 cyst. In the species witli bilateral lophophore, the fibres may be seen towards the apertures 

 of the cells (PI. V, figs. 5, G, i') ; but it is generally impossible, in consequence of the 

 increasing opacity of the superjacent structures, to trace these fibres to any distance posteriorly. 

 In PaludiceUa, whose transparent ectocyst admits of a distinct view of all the contained parts, 

 the fibres are collected into numerous transverse bands (PI. X, figs. 3, 4, v), which we may 

 trace throus;hout the whole extent of the cell. 



