and composed of 20 joints, terminal part biarticulate. Posterior antennae with 

 the penultimate joint scarcely more than half as long as the preceding one; rudi- 

 mentary outer ramus remote from the end of the latter joint, being attached 

 nearly at the middle of its posterior edge. Oral cone extending only to the insertions 

 of the posterior maxillipeds, and gradually tapered. Mandibles with the mas- 

 ticatory part moderately slender, and distinctly denticulated inside the tip; palp 

 with only a single apical seta. Maxillse with the outer lobe a little shorter than 

 the inner, and much narrower, apical seta rather elongated and accompanied by 

 a small hair. Claw of anterior maxillipeds very slender and distinctly subdivided 

 in the middle, outer part evenly curved. Posterior maxillipeds, as usual, more 

 elongated than the anterior ones. Natatory legs of rather feeble structure, the 

 rami, especially in the 4th pair, being unusually narrow. Last pair of legs with 

 the proximal joint well defined from the segment, and produced inside to an 

 acuminate process; distal joint very long and slender, sabre like, anch extending 

 far beyond the middle of the genital segment, outer edge produced, at some 

 distance from the end, to a dentiform projection, the apical part, beyond the 

 process, being provided with 3 or 4 small bristles. 



Male, as usual smaller than female, and of still more slender shape. 

 Anterior antennae composed of 19 joints, the last 2 of which admit of being bent 

 upon the adjoining part. Last pair of legs differing somewhat from those in 

 female, the distal joint being comparatively shorter and almost claviform in 

 shape, and moreover provided with 2 additional blunt spines inside. Genital 

 lobes bipartite. 



Body in both sexes very pellucid and nearly colourless. 



Length of adult female amounting to 0.85 mm., of male to 0.68 mm. 



Remarks. As mentioned above, this form was first described by Brady 

 as a species of his genus Cyclopicera. It was subsequently justly removed from 

 that genus by Canu and placed in the new genus Collocheres established by that 

 author. According to the same author, the Ascomyzon comatulce of Rosoll is 

 identical with the present species. 



Occurrence. I have hitherto only met with this form in a single locality, 

 viz., at Ris0r, south coast of Norway, where a few specimens were taken from a 

 depth of about 30 fathoms among dredged material. Rosoll found this form 

 parasitic on Comatula mediterranea. 



Distribution. British Isles (Brady), coast of France (Canu), Triest (Rosoll), 

 Gulf of Naples (Giesbrecht). 



