506 THE AMERICAN NATURALIST [Vol. XLVII 



crepancy in size, the broad fringe about the body, the 

 large tentacles on either side of the mouth, the absence 

 of muscles of the group type characteristic of the rotifers, 

 and the submarginal anus, seem to negative the idea that 

 the two can be in any way related. 



The medusoid body form, the absence of a protrusible 

 proboscis and the presence of a large branched tentacle 

 on either side of the mouth appear to offer conclusive 

 evidence that Eldonia can not be a "worm." 



The digestive tube of Eldonia resembles that of the 

 heteroradiate echinoderms, and especially that of certain 

 holothurians ; the tentacles on either side of the mouth 

 suggest an affinity with the holothurians; the radial 

 canals, leading to a central ring, are comparable to the 

 radial canals and the central ring of the holothurians; 

 the broad circular muscle about the body suggests a 

 modified longitudinal holothurian muscle, and is of the 

 group type characteristic of the echinoderms ; the broad 

 brim about the body is strikingly similar to the brim 

 developed in certain elpidiid holothurians, such as 

 Euphronides tauueri and Scytoplanes typicus. A pelagic 

 holothurian is known as an inhabitant of the recent seas ; 

 though very different in origin and in affinities from 

 Eldonia, it demonstrates that a pelagic habit is not im- 

 possible in the group. The species of the family Elpidii- 

 dae are preeminently inhabitants of the deep sea; this 

 suggests that the fossil representatives of the family 

 should be found in very early geological formations. 



Therefore Eldonia is a pelagic holothurian, related to 

 the species of the family Elpidiidrc. 



No marine animals are known outside of the holothu- 

 rian family Elpidiida? which have a body form like that 

 of Louisella pedunculata in all its details ; but this spe- 

 cies agrees in every particular with one or other of the 

 species in that family. We can not, therefore, escape the 

 conclusion that Louisella pedunculata should find a place 

 in the family Elpidiida? along with all the recent animals 

 which in any way resemble it. 



