HALOSYDNA GELATINOSA. 387 



extrusion. Both are sharply ridged. A comparatively large number of the preparations, 

 including those from the stomachs of fishes, have the proboscis extruded. 



Only two gastric casca, proportionally short and wide, pass forwards into the peri- 

 pharyngeal space. They are glandular throughout. The tip seems to end in a series of 

 diverging fibrous processes which fix it in the dorso-lateral space. The glandular lining 

 of the stomach is arranged in a series of trumpet-like processes projecting inwards. 



The alimentary canal is generally empty, but in some it contains muddy debris with 

 spicules of sponges and other organic materials, together with the skin, bristles, and 

 hooks of Terebella3. De St. -Joseph says that it feeds on minute crustaceans. Sir J. 

 Daly ell gave a graphic account of its rapacious habits. 



The scales (Plate XXXIII, fig. 12) are eighteen pairs, large — but do not quite cover 

 the dorsum, have a translucent greyish or slightly purplish aspect, or translucent 

 brownish anteriorly, and dull bluish or greyish posteriorly. A pale spot in each marks 

 the scar for attachment. In some of the anterior scales a few dark specks occur along 

 the posterior border. They are soft and more or less irregularly rounded, the first pair 

 having a marked fold at their outer and anterior border. They are devoid of cilia at 

 the margin, and present a minutely cellular aspect by transmitted light. Moreover 

 along the inner and anterior border the surface is densely studded with minute papillae, 

 which are short and broad — generally with a trifid end, two stronger spines laterally, 

 and a smaller and more acute median, as Langerhans pointed out. They extend over a 

 considerable surface of the region indicated. Most of the scales have frills or folds on 

 their outer margin. The finely branched nerves radiate from the scar of attachment 

 throughout the entire scale. They are highly sensitive, so much so that when irritated 

 the annelid will sometimes turn on its back to avoid interference. The scales are fixed 

 to the following bristled feet: 1, 3, 4, 6, 8, and so on to 22, 25, 28, 31, 34, 37, and 38. 



A curious solid rounded body of minute size occurred in the substance of one scale. 

 Its contents were granular. In some also minute opaque white specks were present 

 behind the scar. 



Feet. — The first foot has a spine, the tip of which projects beyond the region in a 

 sheath of cuticle and granular epiderm. No trace of a bristle was present in any 

 specimen examined, and in this it agrees with Malmgrenia. 



In the second foot the dorsal division is marked chiefly by the tip of the spine, 

 which projects a short distance from the dorsum of the long ventral division. Behind 

 the spine is a small tuft of straight, slender, tapering bristles — finely spinous. The 

 combined lobe is comparatively long, nearly cylindrical, and ends in a long conical 

 process for the spine. The ventral bristles are short and slender, extending beyond the 

 conical process at the tip of the foot. They are slightly enlarged at the commencement 

 of the spinous region, and then taper to a hair-like extremity. Their tips, which have a 

 dorsal curve, diminish in length from above downward. The ventral cirrus of this foot 

 extends as far as the tips of the bristles. 



In the third foot the dorsal division is marked only by the projection of the long 

 process enclosing the tip of the spine, and a tuft of similar bristles to those in the 

 previous foot, only the tips are less acute. The ventral division is proportionally 

 shorter and thicker, and the upper border of the process for the spine is now continuous 



