15 



specimens 15 mm. Maximum length of arctic specimens, according to U 

 25 mm. 



Eemarte. In some characteristics, e. g., in the carinated Lack ami 

 the structure of the antennae, this species shows a perplexing resemblance to 

 E. compressa (Goes), and on this account the two species would seem to have 

 been confounded both by Boeck and Bovallius. On closer examination, t In- 

 present species is, however, easily distinguished by the exceedingly elon-alr 

 and slender form of the 3rd pair of pereiopoda, and more especially by the 

 peculiar armature of the propodal joint of these legs. It may be observed 

 that this character is also quite distinct in very young specimens, whereas 

 in these the dorsal processes are very inconspicuous or nearly quite obsolete. 



Occurrence. - Of this form I have taken a few adult and several 

 young specimens at Sorvaer and Hasvig in west Finmark, together with 

 E. compressa and Parathemisto oblivia. 



Distribution. Arctic Ocean: Greenland, Spitsbergen (?); off Nova 

 Scotia (Stebbing). 



Fam. 2. Lycaeidae. 



Cephalon large, subglobular, canaliculated interiorly. Eyes enormous, 

 occupying the entire lateral walls of the head. Antennae originating t'n>m tlie 

 inferior side of the cephalon, very dissimilar; 1st pair rather small, with short 

 uniarticulated peduncle (in female), flagellum rather dissimilar in the two sexes; 

 2nd pair imperfectly developed in female, very large in male and folded up 

 beneath the cephalon. Oral parts more or less imperfectly developed, (inatlm 

 poda simple, or complexly subchelate, The 3 posterior pair of pereiopoda 

 with the basal joint more or less expanded, last pair generally very small. 

 Incubatory lamellae without marginal setae. Pleopoda very powerful. Uropoda 

 normal. Telson triangular. 



Remarks. This family is synonymous with the family Tr//j>//<i i//<l<> 

 of Boeck and Bovallius. I have, however, seen fit to change the name, 

 because, according to the rules of Zoology, the denomination of a family 

 ought to be derived from the genus first established, and as the genus Li/i-<m 

 of Dana undoubtedly belong to the same family as the genus Tr///>//n im of 

 Boeck, the family must of course be named after the former genus, which 

 is by far the older one. According to the restrictions of Mr. Bovallius, the 

 family comprises, besides the 2 above mentioned genera, 5 others, vi/... 



