660 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.39. 



Gray included two species in his genus Nepanthia, 1840 — tessellata 

 (= Ohsetaster longipes), and maculata. No type was indicated; macu- 

 lata, the only available species, is therefore the type. This species 

 has rather long slender rays, and is of the general habit of Ohsetaster 

 longipes.^ I have examined a closely related form from the Philip- 

 pines. The spinelets, though very short, are delicate, and are distally 

 clear and glassy. The middorsal area of externally crescentiform 

 pseudopaxillse is clearly distinguishable from the several dorsolateral 

 longitudinal series of squarish or roundish pseudopaxilhie. Internally 

 both sorts of plates are lobed and imbricated. The crescentic appear- 

 ance of the plates externally is due to the fact that, being arranged 

 in quincunx, two lobes of a plate overlap the abcentral side of the 

 next two plates toward center of disk. In the interval between the 

 lobes the papulae emerge, and the exposed surface has a crescentiform 

 cluster of minute spinelets. The abactinal plates of Hydlinothrix 

 imbxicate in the same fundamental manner, but the plates of theradia] 

 regions have five lobes, so that in overlappmg they frequently but not 

 always join three plates adcentrally. In Hyalinothrix the actinal 

 intermediate plates imbricate to form chevrons, or both longitudinal 

 and transverse series, and the adambulacral spinelets are basally 

 webbed, though not forming fans in the actinal series as in Nepanthia. 

 But the presence of well-developed tabula on the plates of Hyalino- 

 thrix excludes the genus from the Asterinidse (restricted), where the 

 abactinal plates are essentially lamelliform and notched on the 

 adcentral side. 



Neither does it seem possible to place Hyalinothrix in the Chsetaster- 

 idse of Ludwig, for the resemblances are only superficial and are out- 

 weighed by the differences, namely, the small marginal plates, no odd 

 interradial marginals, absence of superambulacrals, the more numer- 

 ous and fundamentally different abactinal plates which lack internal 

 connecting ossicles, the incipiently bilobed ampullse. 



The Ganeriidge have lobed paxilliform plates and small marginals, 

 and the actinal intermediate plates in series extending from the 

 adambulacrals to the marginals. Hyalinothrix, although quite differ- 

 ent from Ganeria, Cycethra, Radiaster, and Scotiaster, appears to belong 

 in this group. Eventually, when the anatomy of the above genera 

 is known, it may be advisable to segregate Hyalinothrix in a special 

 subfamily, the llyalinothricinse. 



HYALINOTHRIX, new genus. 



Type.— Hyalinothrix millespina, new species. 



SteUate, with rounded rays. Marginal plates very small. All plates 

 more or less tabulate, the abactinals strongly four or five lobed, and 

 imbricating by the lobes; abactinal plates numerous, irregularly 



a For figures see Sladen, Challenger Asteroidea, pi. 64, figs. 1 to 4. 



