PALAEODISCUS. 251 



The test is very much flattened in known specimens, as it probably was in Ufe, and is nearly 

 or quite circular in outline. Ambulacra are narrow with two colurtins of low plates, of which 

 about four equal the height of an adambulacral plate. The pore-pairs are about in the middle 

 of the plates and are clearly developed ventrally, as shown in my figures. I cannot understand 

 why Dr. Gregory should say (1897, p. 129) that there are no pores through the plates, 

 but that the podia must have been extended between the sutures. The specimens Dr. Bather 

 kindly loaned me, from the British Museum, have pores distinctly through the plates as in other 

 tj^pical echinoids. Interambulacra with many, apparently about eight or nine columns of 

 highly irregular plates, and, as pointed out by Spencer, the primordial plates are in the basi- 

 coronal row. The plates were evidently extremely thin and bore numerous fine acicular spines. 

 On the peristome there are ten columns of ambulacral plates only. The apical disc is unknown 

 but I think Gregory is not justified in saying probably absent. It is not known to be absent in 

 any echinoid. The lantern, as shown by Sollas, is of the typical character, inclined, with pyra- 

 mids, epiphyses, braces, and even compasses, as figured by Sollas. 



Lower Ludlow, Silurian, near Ludlow, and Church Hill, Leintwardine, England; British 

 Museum, a number of specimens, including those described by Gregory (1897) ; Oxford Museum. 



This species, representing its genus and family, has been much discussed, and on account 

 of imperfections of material, a number of points are still uncertain. Salter referred the type 

 to starfishes, and Zittel (1879, p. 453) followed him. Several authors have considered the 

 type as intermediate between starfishes and echinoids, and as a primitive type. I think recently 

 published material distinctly shows that Palaeodiscus is an echinoid, has no connection with 

 starfishes, and is not a primitive type. 



As regards the ambulacrum, I find no evidence of pores between the plates, as claimed 

 by Gregory, but rather they are through the plates as usual in echinoids. As regards the roofing 

 system of plates described by Sollas and Spencer, they claim that above the typical ambula- 

 cral plates (Sollas, 1899, p. 705, fig. 9) there is a set of narrower plates roofing over the outer 

 . flooring plates. This narrower set is supposed to correspond with the ambulacral plates of a 



/^^U^^'^^^-4\^^a^i¥ehm, superposed on those of a stai'fiA, anotnus the connection of the two groups is 

 supposed to be solved. It is a somewhat novel morphological reasoning from the facts. It 

 would require that we consider the wider or flooring ambulacral plates of Palaeodiscus (which 

 Sollas, 1899, pp. 705, 706, considers the equivalent of the ambulacral plates of later Echini) 

 as equivalent to the adambulacral plates of a starfish grown across the area and meeting in the 

 center line, but I see no basis for such an argument. I think the roofing plates of Sollas are the 

 narrower ambulacral plates of the dorsal side brought in mechanical contact with the wider 

 ventral plates by the flattening of the test, exactly as shown in area B of Hyattechinus rarispinus 

 (Plate 23, fig. 1). The interambulacra are typically echinoid, and from the large number of 

 columns of plates appear to me to indicate a specialized, not a primitive type. Ambulacral 



