CETJSTAGEA. 567 



A specimen -was obtained at Providence Island or He des Eoches, 

 13-20 fms. (No. 183), and also one at Providence Eeef, 24 fms. 

 (No. 215). Both are apparently females. 



The form of the rostrum, together with the armature of the ter- 

 minal segment, distinguish this species from aU with which I am 

 acquainted. 



Kossmann (t. c. zweite Halfte, p. 100) mentions a species of this 

 genus, Oonodactylus hrevisquamatus, Paulson, occurring in the Red 

 Sea, with which 0. elegans may possibly be identical ; but as I have 

 never seen Paulson's work, I can say nothing of the true^ affinities 

 of G. hrevisquamatus. 



There is in the collection a small male from Providence- Island, 

 19 fms. (No. 217), which is allied in many points to the preceding ; 

 but the rostrum is transverse, with its distal extremity deflexed, so 

 that in a dorsal view it appears transvversely oblong, with a straight 

 anterior margin, which does not project beyond the anterior margin 

 of the lateral divisions of the carapace. The penultimate post- 

 abdominal segment is armed with six teeth, including those of the 

 postero-lateral angles. The terminal segment has a smaller longi- 

 tudinal carina on each side of the median longitudinal dorsal carina. 

 The terminal joint of the large raptorial limbs (second maxillipedes) 

 is even less distinctly ventricose at base, and its inner margin is 

 armed with about eight teeth. These characters may be peculiar to 

 the male sex ; but ^f the specimen should prove, on further study, 

 to belong to a distinct species, I would propose to designate it G, 

 hrevirostris (see Plate Lit. iig. C). 



Pseudosquilla empusa (De Haan)* is perhaps the species most 

 nearly allied to our new Gonodactylus ; it has the transverse trun- 

 cated rostrum of the male above described, with the few-spined 

 dactyl of the female ; it is distinguished not merely by the non- 

 ventricose dactyl of the raptorial limbs with its longer spines, but 

 also (if the figure be correct) by the distinctly costated sixth post- 

 abdominal segment, the slightly divergent lateral dorsal carinse of 

 the terminal segment, &c. 



AMPHIPODA. 



1. Moera diversimanus. (Plate LII. fig. D.) 



The body is slender, with the coxse not so deep as their respective 

 segments; the head is about as long as deep, with a small triangular 

 median rostral lobe, and with its antero-lateral angles rounded ; the 

 coxse of the first segment of the body have their antero-lat6ral angles 

 acute and produced below the lateral margins of the head ; in the 

 succeeding pairs the antero-lateral angles are rounded. The first 



* Pauna Japonica, Crust, p. 224, pi. li. fig. 6 (1849). 



