188 A. E. Verrill — Revision Genera and Species of Starfishes. 



of those in the rows next the adambulacral plates have their trans- 

 verse edges bordered by specialized spinules, forming with those of 

 the plates opposed to them, special fascioles [pedicellaires fascio- 

 laires of Perrier), But they are not real pedicellariae. 



True papilliform ]jedicellari{B occur very rarely on the actinal and 

 marginal plates in Paragonaster. 



The jaws are rather large and prominent and bear numerous 

 elongated spinules, both on the actinal surface and the margin. At 

 the oval apex of the jaw there is usually a larger odd median spine, 

 but this is not constantly present in all species, and in certain species 

 it is generally, if not always, lacking. Some specimens may have 

 the odd spine on some of the jaws and lack it on others, so that it 

 cannot be considered of much morphological importance, though its 

 presence is a useful indication of the affinities of certain doubtful 

 species. 



The ambulacral feet have well formed suckers, as in Nymjihaster 

 and Medidster. 



In the general appearance of the abactinal and marginal plates and 

 the granulation of the dorsal surface this group agrees essentially 

 with Mediaster and allied genera. It differs from that group mainly 

 in the more prominent margins of the adambulacral plates, the lack 

 of bivalve pedicellaris, and the more divergent groups of furrow- 

 spinules, together with the usually spinulose covering of the actinal 

 and inferior marginal plates ; but this last character is not constant. 



The singular actinal fascioles are characteristic of many species, 

 but are not constant. The same is true of the odd apical spine of 

 the jaws. 



If true bivalve pedicellarioe were pi-esent we should not hesitate to 

 combine the group with Mediaster and Nyrnphaster in one sub- 

 family. 



On the other hand, the adambulacral plates and spines, the jaws, 

 ;and the form and structure of the dorsal paxillae are very much like 

 those of Plutonaster and allied genera, though the latter does not 

 have the regular serial arrangement of the radial abactinal plates. 

 Nor is there in this gi'oup any distinct arrangement of the actinal 

 plates in rows running from the adambulacral plates to the marginals, 

 as in Plutonaster and allied genera. 



The actinal plates have nearly the same form and are imbricated 

 in the same way as in Mediaster. The adambulacral plates next the 

 dentary plates are somewhat oblique and slightly modified, but much 

 less so than in Plutonaster, and more so than in Tosia and Mediaster .^ 



