136 LIVEEPOOL BIOLOGICAL SOCIETY. 



foot glands. At the apices of the cerata the glands are 

 much more distinctly arranged in ovate or pyriform masses 

 (PI. VII. figs. 7, 8, 9) and there are usually distinct 

 ducts (PI. VII. fig. 9, gV). The cells are smaller, are 

 invariably filled with a clear secretion, and the nucleus 

 is displaced to the side. We find that the cerata are 

 occupied by large blood spaces (the ceratal sinuses, 

 PI. VII. figs. 7 and 8, h.s.) exactly like those of the 

 cerata of Dendronotus arhorescens* 



Ancula is not protectively coloured ; and as it has no 

 cnidophorous sacs, its bright white and yellow colouring 

 and conspicuous appearance on dark rocks seemed for a 

 time inexplicable. From our experiments we have come 

 to the conclusion that it is distasteful to fishes (see below, 

 p. 155), and possibly it is the secretion of these large 

 compound glands at the apices of the cerata w^hich is of an 

 offensive nature. 



In Polycera quadrHineata (PI. VII. figs. 3 and 4) the 

 cerata terminating the lateral ridges on the body, which 

 we regard as representing the cerata of Ancula, contain 

 numerous glands. These are simple pyriform sacs filled 

 with large polygonal granular cells which stain deep 

 crimson with picrocarmine (PL VII. fig. 3, gl). These 

 glands open between the ectoderm cells by long narrow 

 tubular ducts (PL VII. fig. 4). 



In Ancula the large glands in the cerata are somewhat 

 different from those of Polycera quadriUneata. The masses 

 are not so regularly placed and shaped, and the cells are 

 not so granular, but seem to a large extent filled up with 

 a clear secretion, while the nucleus is displaced to one 

 side of the cell. And whereas in Polycera the glands ex- 

 tend nearly all over both sides of the cerata, there being 



* Compare our last report, Proc. Biol. Soc, L'pool, vol. iii., PI. xii. %. 2. 



