24 U.S. NATIONAL MUSEUM BULLETIN 271 



to be a frictional-flotational support for those Amphipoda known to 

 swim in peculiar fashion on their sides. Thus the lateral shield, formed 

 of coxae and articular wings on pereopods, may serve as a kort- 

 nozzle* in reverse by channeling water ahead of the pleopods. 



Good swimmers obviously occur in the Hyperiidea, for they are 

 all pelagonts; nevertheless, reduction of the lateral shield is a major 

 trend within that group. Hence, natatorial correlation to the lateral 

 shield is far from universal and the functional morphologist will 

 find many fruits to pick once he directs his attention to this problem. 

 The loss of pleopods in Amphipoda is a mark of the sedentary life 

 of CaprelHdea, and the terrestrial habits of a few talitroids. Their 

 reduction in ingolfielhds is unexplained but may have some relation- 

 ship to an interstitial or troglobitic life. A few other inquilinous or 

 sedentary gammarideans have reduced pleopods, but they are other- 

 wise remarkably conservative in most domicolous, fossorial, and 

 inquihnous amphipods, presumably because they are often used for 

 creating water currents. 



Function of the last pair of uropods is presumed to be in propulsion 

 and ruddering as in other peracaridans, but uropods 1-2 have seem- 

 ingly little function except as strengtheners for the urosome during 

 explosive flexation that is a part of jumping behavior. Were this 

 the sole function of uropods 1-2 the maintenance of at least a small 

 degree of articular flexibihty of the uropodal peduncles and rami 

 would seem incongruous. Progressive coalescence of uropods 1-2 with 

 the ventral margins of their segments would afford a strong ventral 

 pad to be used as a jumping buffer. Perhaps the small degree of 

 flexibihty provides a better shock absorber than would a soHd uro- 

 some. The increase in ornamentation of uropods on certain fossorial 

 species indicates at least a minimal function in the digging process, 

 but perhaps this ornamentation primarily prevents coarse particles 

 from being lodged in the cracks between uropods and segmental 

 venters. Paddle-like expansion of rami on uropods 1-2 in a few groups 

 (e.g., Synopiidae) points to a swimming function. A few sketchy 

 observations on swimming suggest that the uropods, despite their 

 restricted flexibility, can be laterally splayed and serve in balance or 

 braking during the end of a ghding motion. 



Keduction of uropod 3 and reduction or rigidification of the uro- 

 some occur in various tube dwellers (e.g., Corophiidae), inquihnes 

 (some lysianassids, cressids, thaumatelsonids, possibly kuriids, pro- 

 phhantids, etc.), tunnel makers (Eophhantidae) and some of those 

 amphipods with semipermanent flexion of the abdomen (Phhantidae). 

 Presumably jumping, swimming, and protection from unwanted par- 

 ticles have been reduced in importance in these species while there 



*A tube directing the propeller-wash of ships. 



