GAMMARIDEAN AMPHIPODA 105 



but coxae 3 and 4 are those of a BatTiymedon and not an Aceroides 

 and the mandibuhir palp is as large as that of Bafliymedon. The 

 mandibles of this species bear reseml)lance to those of Arrhis phijllonyx 

 (M. Sars) (see G.O. Sars, 1895, pi. 119) but the well-developed rostrum 

 and the short peduncular articles of antenna 1, plus the coxae, are 

 distinguishing characters of Bathymedon. 



The long rostnim differentiates B. ca'mo from B. candldus J. L. 

 Barnard (1961), B. cov'dhani J, L. Barnard (1961) and B. paljjolis 

 K. H. Barnard (1916), all of which have gnathopods faintly resem- 

 bling those at hand. The greatly elongated gnathopod 2 of B. saus- 

 sui'ei (Boeck) (see Sars, 1895, pi. 118, fig. 1) distinguishes it from B. 

 camo. The posterior lobe of gnathopod 2, article 5, distinguishes the 

 new species from others of Batliyrrhedon having a large rostrum, such 

 as B. Jongimanus (Boeck) , B. acAitlfrons Bonnier, B. subcarinntus Bul- 

 ycheva (1952), B. roquedo J. L. Barnard (1962d), B. nansenl Gurjan- 

 ova (sec 1951) and^. 7aw^.s<for^Gurjanova (1951). 



The sharp posteroventral lobe of coxa 4 is unusual for the genus and 

 suggests the possibility that, as an intragcneric character in Batliy- 

 rnedon^ its use as a generic character in Paroedlceroides Schellenberg 

 (1931) and Oedicerojisls Liljeborg is erroneous. 



A male fragment of this species from 7358 has antenna 1 as long 

 as that of B. Jongimanus and extending as far beyond the peduncle of 

 antenna 2 as in that species but antenna 2 was only just commencing 

 the male elongation at the time of death; the rostrum is slightly 

 smaller than that of the female and the anterior edge of the lateral 

 cephalic lobe is slightly more concave. 



Bathymedon candidus J. L. Barnard 



Figure 48 

 Bathymedon candidus J. L. Barnard, 1961, pp. 84-8."5, fig. 52. 



Although the head is better defined in the material at hand and hence 

 appears slightly different than the holotype (Makassar Strait), no 

 other major differences are apparent and the specimens are identified 

 with B. candidus, despite the great geographical separation of the two 

 known collections. 



Hidden among the anterodistal setae of article 5 on gnathopod 1 are 

 several small, stout spines, not present on gnathopod 2. Their occur- 

 rence on the holotype has not been ascertained. 



Materdvl.— Station 7231 (14). 



Distribution. — Makassar Strait and middle Baja California, 2000- 

 2398 m. 



