OBSERVATIONS ON POLYZOA. 227 



drawn in, as they often are, to less than one third of their 

 full length, I have ventured to assume that it also exists 

 there (PI. 11, fig. 1, E"'). 



(4) The fourth or epithelial membrane, lines the interi- 

 or, investing all the muscles and the digestive system. It 

 is ciliated upon the perigastric region, and upon the in- 

 terior of the arms and lophophore, but not in the tenta- 

 cles or upon the alimentary canal. On the abdominal side, 

 a double layer, or fold, of this membrane, which I have 

 named the Brachial Collar, constitutes a partial diaphragm 

 reaching about half way round the oesophagus. On the 

 dorsal side it is disconnected from the lophophore, and 

 hangs into the perigastric space, partitioning off the inside 

 of the epistome, and a space below in which the ganglion 

 is suspended. There are numerous fibres upon the inner 

 side of this diaphragm attached to the oesophagus and en- 

 docyst, between the bases of the arms, having sufficient 

 contractile power to deeply infold that part of the body 

 wall. 



Prof. Allman mentions but two membranes in the en- 

 docyst, one, an outer large celled layer, equivalent to my 

 first and second membranes, and another, an inner layer, 

 equivalent to my third and fourth membranes. Through- 

 out its whole extent, the fourth or epithelial layer is lined by 

 muscular fibres. These cross the transverse fibres of the 

 third layer at right angles (fig. 4, E""). and both were re- 

 garded by Prof. Allman as a single inner layer of reticula- 

 ted muscles. The longitudinal fibres, however, are inva- 

 riably next to the fourth membrane, and remain attached 

 to it, whenever, as in the neural diaphragm, it parts from 

 the other layers. The transverse fibres, also, never seem 

 to be connected with the longitudinal, wherever a good 

 definition of either has been obtained. No transverse 

 fibres are visible on the neural diaphragm; and on the in- 

 vaginated fold (Fig. 8), and the oesophagus (PL 11), no 

 longitudinal fibres are visible.* In the two latter they are 



*Xote. Since the printing of the plates, I have, in reviewing these 

 pages, changed my opinion and now estimate the longitudinal fibres, as 

 of equal importance with the transverse, and consider them a fourth 

 ]ayer of muscular fibre, the epithelial becoming a fifth membranous lay- 



