34 NEW YORK STATE MUSEUM 



more prominent than those surrounding it, and just inside of the 

 adjoining proximal supramarginals is a small but distinct plate 

 interradial in position" (Schuchert). In the Saugerties specimens 

 the number of accessory plates in this area is easily but half that 

 of Hall's types and since the accessory plates are also relatively 

 smaller the central plate and the ring of five plates in this area 

 are very much more prominent. In a young specimen (length of 

 ray but 12 mm) this central plate and the surrounding ring of five 

 plates, have already the size of those in the mature and gerontic 

 specimens, and since there are but very few small accessory plates 

 between them, they are nearly in contact. Outside of this ring there 

 follows almost immediately a closed ring of ten equal, subcircular 

 adjoining plates. We have here, therefore, a condition of the disk 

 corresponding to that of the Lower Devonian Spaniaster; but we 

 have a still smaller specimen and this leaves no doubt that originally 

 the disk had no accessory plates at all. The- central plate and the 

 ring of small plates which in the mature disk are hardly dis- 

 tinguishable from the accessory plates and have been counted with 

 them, are hence primary plates and homologous to the centrodorsals 

 and the " underbasals " respectively. The next ring of ten plates 

 consists of the proximal radial plates and the interradial plates. 

 That these are yet in contact, subcircular and equal in shape, is 

 also a primitive character. It seems therefore that we have in these 

 young still the primary skeleton of the Phanerozonian type, namely, 

 the centrodorsal, a first ring of five primordial radials, and a second 

 ring of ten plates, five of which are the second radials, and five 

 interradial pieces, the primordial supramarginals. This primitive 

 skeleton is still seen in the Ordovician starfish, Hudsonaster, where, 

 however, also accessory pieces have already been inserted between 

 the centro-dorsal and the first ring of primary radials. 



The abactinal side of the arms has also an entirely different 

 aspect in the gerontic types from those obtained at Saugerties. In 

 the former the radials and supramarginals are floating, so to say, 

 in a mass of minute accessory plates which fill all interspaces. In 

 the Saugerties specimens only small groups of minute accessory 

 plates are developed in the corners where two adambulacrals and 

 supramarginals come in contact in the proximal portion of the rays. 

 They are so little prominent that they were at first overlooked. In 

 a few cases they unite to a thin line between the adambulacrals and 

 supramarginals in the proximal portion of the arms. Nevertheless 

 they appear so early that already in the half-grown individuals 

 single accessory plates are seen in the proximal corners (see plate 



