461 



outer jiai't of the Haidaiigerfjoid, at Siiiulc and Mosterliavn; at Cliiistiau- 

 sund; and in the Trondhjemsfjord. In all 3 localities it occurred quite solitary 

 in a very considerable depth, ranging from 150 to 400 fathoms. 

 Distribution. — Greenland (Hansen). 



Fam. 19. Atylidae. 



Body generally not very slender, with the back, as a rule, much 

 curved. Urosome comparatively strongly built, and having the last 2 segments 

 completely fused together. Cephalon rather deep, and more or less produced 

 in front, lateral corners small or obsolete. Coxal plates of moderate size or 

 very small, 5th pair with the anterior lobe deeper than the posterior. Eyes 

 distinct. Antennge slender, and more or less distinctly modified in the male, 

 the superior ones, as a rule, without any accessory appendage. Oral parts 

 well developed. Mandibles short and compact, with the molar expansion very 

 massive, palp present or wanting. Maxillipeds with the masticatory loljes 

 generally very large, palps comparatively small. Gnathopoda not very 

 strong, subcheliform, and somewhat unequal, the posterior ones being generally 

 more slender than the anterior. Pereiopoda not much elongated, the 3 

 posterior pairs generally strongly reflexed, and having the dactylus in some 

 of them inverted. Branchial lamellae large and often of complicated structure, 

 being folded transversel3^ Penultimate pair of uropoda much smaller than 

 the 1st pair; last pair reaching generally beyond the others, and having the 

 basal part short, and the rami subequal. Telson more or less developed, and 

 deeply cleft. 



Ecmarls. — In the restriction here adopted, this family comprises 

 but a small number of the forms referred by Boeck to his subfamily Atylincc. 

 On the other hand, I have felt justified in including in the same family the 

 genera upon which Boeck founded his subfamily Bexaminince. True, in the latter 

 the matidibular palp is wanting, l)ut I cannot regard this characteristic in itself 

 as being sufficient to warrant such a separation, since there are other families 

 that include forms both with and without mandibular palps, for instance the 

 family Stenothoidce. In other respects, there is evidently a very close relationship 

 between the Dexaminina and the forms of the Atylinm included in the present 

 family, both as to outward appearance and to most of the structural details. 

 As a most curious feature, common at least to all the Norwegian forms of 

 the family, may be named the complete fusion of the 2 last segments of the 



