32 U.S. NATIONAL MUSEUM BULLETIN 271 



The major line of evolution out of the Gammaridae appears to pass 

 through eusirid organisms in the next block below; this trend is first 

 marked by reduction or loss of accessory flagellum and radiates further 

 by various modifications to be discussed in a later paragraph. The 

 point to be made is that the change from gammarids to eusirids 

 seems to be less revolutionary than the changes from gammarids to 

 various famihes kept mthin the gammarid block, and each of those 

 probably should be accorded the rank of a block. The mark of their 

 relationship to gammarids is the retention of the accessory flagellum 

 in multiarticulate condition. A presumption that an accessory 

 flagellum could be redeveloped once lost would have to be invoked 

 in order to bring some of those gammarid-like families down into the 

 eusirid block where grades of other structures show some resemblance. 

 I see Httle objection and a slight precedent to that possibility in the 

 fact that many Gammaridea add articles to the accessory flagellum 

 with body growth; is there a strong difference in metameric potential 

 between 1 and 2 or zero and 1?. If all the gammarid-block families 

 were placed in the eusirid block only two, Liljeborgiidae and Vitjazi- 

 anidae, could be further drawn from eusirids by reestabhshing the 

 accessory flagellum. There is no logic in funnehng all other gamma- 

 rideans through a eusirid block when many of them can be extended 

 directly from the Gammaridae. Gammaridae are no less diverse than 

 the three major families of the eusirid block. Each of the gammarid- 

 block families shows some tendency in their advanced genera towards 

 reduction in the accessory flagellum, thereby suggesting that the 

 presence of the ramus in those families is primordial. 



Melphidippidae have the structure of gammarids in which the 

 coxae have become shortened, and the gnathopods enfeebled hke 

 females of the gammarid genus Cheirocratus, plus an elongation of 

 uropod 3. CephaHc ocular bulges and a strong trend towards reduction 

 of the accessory flagellum are characteristic. 



Vitjazianidae have the general appearance and ecology of some 

 pelagic eusirids but their retention of a 3-articulate accessory flagellum 

 and the conjoint base of the primary flagellum indicate a mode of 

 evolution different from that of eusirids and thus the Vitjazianidae 

 are maintained in the gammarid famihes even though they may be 

 more advanced than eusirids or gammarids. Gnathopod 1 has become 

 simple and the most specialized vitjazianids have extremely reduced 

 coxae. The mouthparts maintain a basic gammarid structiu-e. 



Hyperiopsidae form another pelagic theme. The typical genus 

 resembles hyperiid Amphipoda but retains the maxillipedal palps; 

 the second genus, Parargissa, differs so remarkably in its overall 

 appearance that it is debatably an hyperiopsid but mouthparts, 

 antennae, and pereopods seem to confirm the relationships between 



