26 U.S. NATIONAL MUSEUM BULLETIN 271 



dorsal shield as a major, albeit rare, adaptation of Amphipoda. 

 Dorsal shielding is also seen in the calliopiid Amphithopsis and it 

 thereby forms a strong link among other Calliopiidae, the two 

 quixotic genera, and is conducted into the Laphystiopsidae. l^o 

 some extent body depression also occurs in a few Podoceridae and 

 this seems significant in light of their presumed derivation from 

 tubicolous isaeids having lost the tube-spinning glands. This loss 

 may be correlated mth such dorsal shielding because both methods 

 of protection would seem unnecessary together. 



Temnophlias, a "phhantid," has also been included by Gurjanova 

 as a member of the progression from Amphithopsis through Sancho 

 into the Phliantidae but another suggestion might be made: that 

 Temnophlias is really a cyUndrical organism with secondary pleuroni- 

 zation of the pereonites similar to munnid isopods. It may have 

 affinities with the EophUantidae, the most strongly developed of the 

 cyUndrical gammarideans. Cylindricality and dorsal shielding are 

 difl&cult to separate as various corophiids, chelurids, and aorids have 

 always been considered to be depressed organisms rather than 

 cyhndrioid. Both terms partially apply to these groups. But one 

 may consider that those organisms have substituted domiciliary 

 habits in the form of tube building or burrowing for the true lateral 

 shield and that dorsal depression is a consequence of cylindricaliza- 

 tion that cannot be carried too far without rendering the organism 

 positionally unstable. Some depression is required for the organism 

 to maintain a crawUng equilibrium. Gurjanova appears to regard 

 Phhantidae (dorsal shields) and Eophhantidae as cohesive, but 

 eophhantids are strikingly cyUndrical. There is now evidence that 

 eophUantids are phycophilous burrowers and this would correspond 

 with the habitats of other famiUes Uving in diverse kinds of "tunnels." 

 The spherical heads and cyUndrical, rotatable necks of eophUantids 

 are suggestive of Umnoriid isopods and are presumably associated 

 with the tunneUng habit. They strongly contrast with PhUantidae 

 morphologicaUy but there is one small difficulty in completely segre- 

 gating the two famiUes in that one phUantid is known to be a ligni- 

 vore; one must presume that phUantids are not tunneUng Ugnivores 

 on morphological evidence alone. 



The cyUndrioid Colomastigidae again may be thought of as tun- 

 nelers or domiciUaries because they probably inhabit tests or tissues 

 of sessile invertebrates. But inquiUnous behavior is scarcely confined 

 to cyUndrioid amphipods, for anamixids and dexaminids, both with 

 strong lateral shields, are undoubted inquiUnes because of their 

 mouthparts (Anamixidae) or their known ascidiophilous behavior 

 {Polycheria, Dexaminidae). Piercing and sucking mouthparts are 

 common in many other gammarideans such as Acanthonotozomatidae 

 and various Lysianassidae. 



