A NEW TYPE or AMPIIIPODA. 131 



The two anterior pairs of uropods normal; the third pair 

 reduced. The telson thick, ronnded. 



"When more material of the two species established in this 

 paper has been procured, and new forms have been detected, it 

 is possible that one or a few characters (marsupial plates, &c.) 

 may be added to this diagnosis, and that some of the characters 

 enumerated will be found to be only of generic value. From 

 the structure of the antennulge and antemise, of the prehensile 

 hands, the uropods, &c., it would be possible to compose 'a 

 diagnosis for the genus, but only two species being as yet known 

 a diagnosis would be rather uncertain, and I abstain therefore 

 from the attempt. 



IV. The Hank and Position of tlie Fa»2?7?/ Ingolfiellidge 

 in the System. 



The order Amphipoda is generally divided into three tribes 

 or suborders — Hyperina (Hyperidea), Gammarina (Gammaridea), 

 and Caprellina (Caprellidea, Lsemodipoda) ; and this division is 

 maintained by the best modern authorities (Stebbing, G. O. 

 Sars). The Hyperina is, so far as I can see, a very natural 

 group compared with the other two tribes taken together, even 

 if the diagnosi.s hitherto given for it should prove capable of 

 being a little altered. The separation of the Caprellina from 

 the Gammarina is, in my opinion, not very important from a 

 strictly scientific point of view, but it is certainly very practical, 

 and may therefore be accepted. If this separation is maintained, 

 I think it necessary to establish a fourth tribe for the reception 

 of the Ingolfiellidse : this family is, according to the characters 

 pointed out above, even more distant from the Corophiidse, 

 Dulichiidae, &c., or from the Caprellidse, than the last-named 

 family, containing such genera as Cercops and Proto, is remote 

 from the Dulichiidse. So long as forms intermediate between 

 Ingolfiella and some genus of the tribe Gammarina are not 

 discovered, it will be necessary to maintain the Ingolfiellidse as 

 a tiibe of the same rank as the Caprellina, and consider both 

 2>,^ parallel and analogous; if the last-named tribe is withdrawn 

 the tribe lugolfiellkia must also be suppressed and reduced to 

 only a family of the Gammarina, which then should contain all 

 genera not belonging to the Hyperina. 



