REVISION OF PALEOZOIC STELLEROIDEA. 273 



wide groove, but farther on appear to come in contact on their inner 

 margins. These plates consist of joints closely articulated together, 

 with a longitudinal foramen: their lower sides are marked lengthwise 

 by a comparatively wide but not deep groove. The body of the plate 

 has, on each side, a lateral arching process which is jointed at the 

 two extremities, and separated in the center by a pore-like perfo- 

 ration. There are ten oral plates, two from each division of the ray: 

 these plates are expanded vertically; their extreme points have the 

 inner edges slightly curving; the lower external faces are slightly 

 indented, or crenulate ; the surface of attachment is wide and strong, 

 and constricted at the base by a distinct groove, beyond which it 

 again expands. 



"This form, if really without a disk, differs essentially from Prot- 

 aster; and there are also other differences, which may make it 

 necessary to constitute a distinct genus when better specimens shall 

 be obtained." 



Formation and locality. In the Burlington limestone at Burlington, 

 Iowa. There are three specimens of this species in the Wachsmuth 

 collection (Nos. 10, 11, and 12) at Harvard University. 



ONYCHASTER CONFRAGOSUS Miller. 



Onychaster confragosus MILLER, Seventeenth Rep. Geol. Surv. Indiana, 1892, 

 p. 684, pi. 12, figs. 6, 7; advance extras, 1891, p. 74, pi. 12, figs. 6, 7. 



Original description. "This species is so different from the one ^ast 

 described [0. asper], and also from the type of this genus, that it may 

 not be congeneric with either of them. I have no doubt that it 

 belongs to the same family. I have seen only the dorsal side of the 

 disk and the dorsal and lateral sides of the rays. There is no such 

 integument of small plates covering the dorsal side as belongs to 

 0. asper, but the surface is covered with short spines. 



"The central part of the disk is somewhat injured in our specimens, 

 and its character can not be definitely determined. It is surrounded 

 by large, centrally convex, six-rayed plates. These rays seem to be 

 the elevated ridges that separate the sockets for the insertion of the 

 spines. Outside of this circle of ten six-rayed plates, there is a circle 

 of quadrilateral plates more or less sculptured by the sockets for the 

 spines, and here the radials may be said to commence. The dorsal 

 side of each ray consists of three series of plates, and there is one 

 series upon each side, or five series in an arm. The first plates are 

 connected laterally by smaller ones, in the angular depressions, 

 between the commencement of the rays, which form part of the disk* 

 All of the radial plates are more or less sculptured by the depressions 

 for the insertion of the spines. The spines are longer than the diam- 

 eter of a plate, bulbous at the lower end, and taper to an obtuse point. 

 The arms are long, round on the dorsal side, flexible and capable of 

 being rolled up on the ventral side or being twisted laterally. No 



