MARINiE GAMMARIDEAN AMPHIPODA 39 



include some ischyrocerids also. The Ampithoidae are fairly uniform 

 by virtue of their third uropods and the Ischyroceridae, though ana^ 

 ogous to ampithoids because of the development of a rudimentarily 

 uncinate condition on the outer rami, are recognizable by the elonga- 

 tion of the peduncle. Ischyroceridae rarely have the reversed gnatho- 

 podal domination. One corophiid, Ericthonius appears to combine 

 features of Corophiidae, Aoridae, and Ischyroceridae, but it is not the 

 only "isaeid" genus that gives trouble to the systematist. 



The lignivorous Cheluridae have numerous resemblances to the 

 isaeid-corophiid line, but the fully pygidized urosome reveals suture 

 marks indicating that urosomite 3 has become extremely enlarged, a 

 feature unique to this famUy. There is no better demonstration of the 

 extreme genetic potential of numerous Amphipoda in replicating long 

 lost structures than by considering the enormous, Maera-like 

 gnathopod 1 of Chelura insulae Caiman. 



Families without clear relationships to others are the Lysianassidae, 

 Phhantidae, Colomastigidae, AmpeUscidae, and potential families are 

 represented by Didymocheila, Ceina, and Biancolina. 



The Ampeliscidae are a very advanced group, apparently completely 

 divorced from the isaeid complex, which have developed pereopodal 

 glands and spinning tubes of a different form from those of isaeids. 

 Ampeliscid morphofunction is also discrete, as far as we know. 

 Ampeliscids have a mixture of characters represented by argissids and 

 atylids but those two families otherwise bear no relationships. Thus, 

 the line on graph 1 connecting Ampeliscidae to Argissidae and Atylidae 

 represents only a focus on structural grades. 



Despite similarities to the Sancho-Chosroes members of the Calliopi- 

 idae, the Phliantidae bear relationships to certain dorsoventrally de- 

 pressed podocerids, like Podocerus and I am inclined to the view that 

 the latter suggestion has great merit and should be examined carefully. 

 Phliantids also have about as much in common with Laphystiopsidae 

 (mostly dorsoventral depression) as they do to Sancho and Chosroes. 

 Apical curls on brood lamellar setae plus many similarities in buccal 

 and urosomal parts suggest phhantid affinities with Talitroidea. 



Relationships of the Colomastigidae are difficult to trace. The body 

 form is subcylindrical but the head is not of the spheroid kind found in 

 the EophUantidae, the rami of uropod 3 are present and elongate, the 

 peduncle is elongate and the mandible apparently lacks a true incisor, 

 that characteristic having been replaced by an enlarged spine row. 

 Affinities with leucothoids are seen in maxillipeds, uropods, and telson. 



The Lysianassidae are almost as fully diverse as all of the other 

 Gammaridea put together and comprise about 20 percent of all marine 

 genera and species of the suborder. They are united together by their 

 clearly recognizable gnathopod 2 and furthermore share an almost 



