94 U.S. NATIONAL MUSEUM BULLETIN 2 60 



Arrliis Stebbing, with its type-species A. 'phyUonyx (M. Sars) (see 

 G. O. Sars, 1895) and its congener^, luthkei Gurjanova (1936, 1951) 

 may be eliminated from the discussion as a cohesive pair distinguished 

 by the westwoodilla mandible having an essentially untoothed, poorly 

 projecting cutting edge. Arrhis completely lacks a rostrum, has an 

 elongated antenna 1 with long article 2, equal to {A. luthkei) or much 

 longer than {A. phyllonyx) article 1 and combines these characters 

 with normal oedicerotid pereopods 1 and 2 and subchelate gnathopods 

 having posterior lobes on the fifth articles that either do (on gnathopod 

 2) or do not (on gnathopod 1) guard the sixth articles. 



Aceroides Sars (1895) was erected with special reference about simi- 

 larities to Halicreion Boeck, a genus distinct from those under dis- 

 cussion because of the elongated uropod 3 ; but Halicreion must be kept 

 in mind because uropod 3 often is lost in specimens being described 

 as new species. Sars also wrote remarks on Arrhis (then known as 

 Aceros) which established such characters of generic importance as 

 differences in mandibles, lengths of first antennae and their articles, 

 and rostra. To a large extent, however, rostra and antennae have 

 been used as secondary generic characters in subsequent years, e.g., 

 Oediceroides antenn,atus K. H. Barnard (1937), which was accepted 

 into Oediceroides even though it has elongated articles of antenna 1. 

 The plethora of variations in rostra of Monoculodes^ Westwoodilla^ 

 Bathymedon^ and Oediceroides apparently has been so overwhelming 

 as to be ignored at the generic level. Sars also was aware of differ- 

 ences in first antennae in his erection of Monoculopsis and its distinc- 

 tion from Monoculodes. 



The extreme reduction in the rostrum of Arrhis is realized in deep- 

 sea species discovered in the last few decades, such as the type-species 

 Anoediceros hanseni Virlot (1932), Oediceroides (Patoides) synparis, 

 0. (Lopiceros) forensia^ several species of Bathymedon^ Aceroides li- 

 micola^ and several new species described herein. These taxa also 

 lack eyes, and specimens often have badly damaged uropods and an- 

 tennae. Therefore, they are difficult to classify because many shallow- 

 water genera are distinguished by ocular structures and a few by the 

 length of uropod 3. 



K. H. Barnard (1925, p. 350) complained of the large number of 

 monotypic oedicerotid genera in assigning his A. limicola to Aceroides 

 but he might have been justified in erecting a new genus on the basis of 

 the strongly projecting mandibular cutting edge. This edge is short 

 and blunt in the type-species, Aceroides latipes (Sars) (see 1895). 



All species of Monoculodes have a well developed rostrum strongly 

 projecting beyond the lateral cephalic lobes; it is usually longer than 

 half the length of article 1 of antemia 1. Most species of Monocu- 



